res_analysis.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "# Analysis Overview of Advanced RAG Techniques\n",
        "\n",
        "This notebook serves as a detailed analytical resource within a larger experimental study on advanced Retrieval-Augmented Generation (RAG) techniques, as detailed in our repository. It utilizes experimental outputs generated by tonic_main.py to analyze the performance and implications of these techniques.\n",
        "\n",
        "## Highlights:\n",
        "\n",
        "- **Data Insight**: Initial processing and visualization of experimental data provide a foundation for understanding the performance landscape of various RAG techniques.\n",
        "- **Statistical Validation**: We employ ANOVA and Tukey's HSD tests to statistically evaluate the performance differences across techniques, ensuring our conclusions are robust and reliable.\n",
        "- **Focused Analysis**: While encompassing a broad spectrum of RAG techniques, specific attention is given to high-impact findings and their significance within the larger context of LLM enhancement.\n",
        "\n"
      ],
      "metadata": {
        "id": "BsrbcXm377DU"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Data preparation"
      ],
      "metadata": {
        "id": "wREKLdd771IA"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "\n",
        "# Load the experimental data from an Excel file\n",
        "df = pd.read_excel('checkpoint_2.xlsx')\n"
      ],
      "metadata": {
        "id": "fDTTBdETIrJj"
      },
      "execution_count": 22,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import ast\n",
        "\n",
        "# Parse the 'OverallScores' column to convert string representations of dictionaries into actual dictionaries\n",
        "try:\n",
        "    df['OverallScores'] = df['OverallScores'].apply(ast.literal_eval)\n",
        "except ValueError as e:\n",
        "    print(f\"Error encountered: {e}\")\n",
        "\n",
        "# Extract these dictionaries into separate columns and combine with the original DataFrame\n",
        "metrics_df = df['OverallScores'].apply(pd.Series)\n",
        "expanded_data = pd.concat([df.drop(columns=['OverallScores']), metrics_df], axis=1)\n",
        "\n",
        "# Display the first few rows to verify the transformation\n",
        "expanded_data.head()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "SOdtu4IySIRC",
        "outputId": "fdff2504-5892-4732-d85a-6e429d8d42ed"
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   Unnamed: 0  Run               Experiment  retrieval_precision  \\\n",
              "0           0    1  Classic VDB + Naive RAG             0.697819   \n",
              "1           1    2  Classic VDB + Naive RAG             0.669782   \n",
              "2           2    3  Classic VDB + Naive RAG             0.704050   \n",
              "3           3    4  Classic VDB + Naive RAG             0.719626   \n",
              "4           4    5  Classic VDB + Naive RAG             0.700935   \n",
              "\n",
              "   answer_similarity  \n",
              "0           3.046729  \n",
              "1           3.084112  \n",
              "2           3.158879  \n",
              "3           3.102804  \n",
              "4           3.140187  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-13366d4b-8918-4b81-9777-1e2b76f909fa\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Unnamed: 0</th>\n",
              "      <th>Run</th>\n",
              "      <th>Experiment</th>\n",
              "      <th>retrieval_precision</th>\n",
              "      <th>answer_similarity</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>0.697819</td>\n",
              "      <td>3.046729</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>0.669782</td>\n",
              "      <td>3.084112</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>0.704050</td>\n",
              "      <td>3.158879</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>0.719626</td>\n",
              "      <td>3.102804</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>5</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>0.700935</td>\n",
              "      <td>3.140187</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-13366d4b-8918-4b81-9777-1e2b76f909fa')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-13366d4b-8918-4b81-9777-1e2b76f909fa button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-13366d4b-8918-4b81-9777-1e2b76f909fa');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-88b02571-4639-4117-8d9a-ced1073f971c\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-88b02571-4639-4117-8d9a-ced1073f971c')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-88b02571-4639-4117-8d9a-ced1073f971c button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "expanded_data",
              "summary": "{\n  \"name\": \"expanded_data\",\n  \"rows\": 160,\n  \"fields\": [\n    {\n      \"column\": \"Unnamed: 0\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 46,\n        \"min\": 0,\n        \"max\": 159,\n        \"num_unique_values\": 160,\n        \"samples\": [\n          105,\n          108,\n          141\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Run\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 2,\n        \"min\": 1,\n        \"max\": 10,\n        \"num_unique_values\": 10,\n        \"samples\": [\n          9,\n          2,\n          6\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Experiment\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 16,\n        \"samples\": [\n          \"Classic VDB + Naive RAG\",\n          \"Classic VDB + Cohere Rerank\",\n          \"Classic VDB + HyDE + LLM Rerank\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"retrieval_precision\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.07553906603369674,\n        \"min\": 0.6168224299065419,\n        \"max\": 0.9127725856697819,\n        \"num_unique_values\": 121,\n        \"samples\": [\n          0.7672955974842768,\n          0.8068535825545171,\n          0.7009345794392524\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"answer_similarity\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.12195963844608301,\n        \"min\": 2.7570093457943927,\n        \"max\": 3.2710280373831777,\n        \"num_unique_values\": 47,\n        \"samples\": [\n          2.8598130841121496,\n          2.803738317757009,\n          2.94392523364486\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Visualization"
      ],
      "metadata": {
        "id": "GIdJ8req8ebN"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Boxplot of Retrieval Precision"
      ],
      "metadata": {
        "id": "QKxVZVVA8UVQ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "\n",
        "# Visualization of Retrieval Precision across different experiments\n",
        "plt.figure(figsize=(10, 8))\n",
        "sns.boxplot(x='retrieval_precision', y='Experiment', data=expanded_data, palette=\"Set3\", orient='h')\n",
        "plt.title('Boxplot of Retrieval Precision by Experiment')\n",
        "plt.xlabel('Retrieval Precision')\n",
        "plt.ylabel('Experiment')\n",
        "plt.tight_layout()\n",
        "plt.show()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 918
        },
        "id": "kCPEH0FIf3Xn",
        "outputId": "13bcfb5e-bdc0-4777-d61d-ca170c29830c"
      },
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-24-9398e5d2243c>:6: FutureWarning: \n",
            "\n",
            "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `y` variable to `hue` and set `legend=False` for the same effect.\n",
            "\n",
            "  sns.boxplot(x='retrieval_precision', y='Experiment', data=expanded_data, palette=\"Set3\", orient='h')\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x800 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Boxplot of Answer Similarity"
      ],
      "metadata": {
        "id": "sapUpLXo8XHs"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Visualization of Answer Similarity across different experiments\n",
        "plt.figure(figsize=(10, 8))\n",
        "sns.boxplot(x='answer_similarity', y='Experiment', data=expanded_data, palette=\"Set3\", orient='h')\n",
        "plt.title('Boxplot of Answer Similarity by Experiment')\n",
        "plt.xlabel('Answer Similarity')\n",
        "plt.ylabel('Experiment')\n",
        "plt.tight_layout()\n",
        "plt.show()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 918
        },
        "id": "GLR15z_1K88g",
        "outputId": "5c0d2139-718e-4456-b7ea-e24aa0fe95e1"
      },
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-25-839d46f944a9>:3: FutureWarning: \n",
            "\n",
            "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `y` variable to `hue` and set `legend=False` for the same effect.\n",
            "\n",
            "  sns.boxplot(x='answer_similarity', y='Experiment', data=expanded_data, palette=\"Set3\", orient='h')\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x800 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA94AAAMWCAYAAAAH1l7yAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOzdeVhU5fs/8PewjYDD6gIoigiDiBviliZCUqiBG2IiuYGagpLgglgmbpErSa5RgGtoLtVHXEjBMJfSDNPUFBRJhSxkVWQ9vz/8cb6MAwrqgMr7dV1cMc927ufMweaec85zJIIgCCAiIiIiIiIilVCr7wCIiIiIiIiIXmdMvImIiIiIiIhUiIk3ERERERERkQox8SYiIiIiIiJSISbeRERERERERCrExJuIiIiIiIhIhZh4ExEREREREakQE28iIiIiIiIiFWLiTURERERERKRCTLyJiIjopSaRSBAaGlrfYSg4c+YMevfuDV1dXUgkEiQnJ9d3SK+NtLQ0SCQSxMTEvLAxjx07BolEgmPHjoll48ePh4WFxQvbRoW6Ol4tLCzg5uam8u28yqp634nqCxNvIiKiBiomJgYSiUThp1mzZnB2dsbBgwfrO7zndunSJYSGhiItLe2FjltSUgJPT0/cu3cP4eHh2Lp1K1q3bv3UfgcOHIBEIoGZmRnKy8tfaEyviv/973/o168fmjVrBh0dHVhaWmLkyJE4dOhQfYemMidPnkRoaChycnLqO5Raq/gSpLqfzz77rL5DfC3s2LEDn3/+eX2HQSqmUd8BEBERUf1atGgR2rRpA0EQ8M8//yAmJgaDBg3C//73v1f6jNqlS5ewcOFCODk5vdAzm6mpqbh58yYiIyMxceLEGvfbvn07LCwskJaWhoSEBLi4uLywmF4FK1euxOzZs9GvXz+EhIRAR0cHKSkpOHLkCGJjYzFgwAAAQOvWrVFYWAhNTc0Xtm1HR0cUFhZCS0vrhY1ZncLCQmho/N9H7JMnT2LhwoUYP348DAwMVL59VfDy8sKgQYOUyu3t7eshmpqry/f9eezYsQMXL17EjBkz6jsUUiEm3kRERA3cwIED0a1bN/G1r68vmjdvjm+++eaVTrxV5e7duwBQqyTq/v37+P777xEWFobo6Ghs3779tUu8BUHAw4cPoa2trVRXWlqKxYsX4+2330Z8fLxSfcU+BR5dqt2oUaMXGpuamtoLH7Oy8vJyFBcXo1GjRirdTn3p2rUr3n///foOo8YePnwILS0tlb/vRLXBS82JiIhIgYGBAbS1tRXO2gGPkseZM2fC3NwcUqkUNjY2WLlyJQRBAPDoTF+7du3Qrl07FBYWiv3u3bsHU1NT9O7dG2VlZQAe3V/buHFjXL9+Ha6urtDV1YWZmRkWLVokjvckv//+OwYOHAg9PT00btwY/fv3x+nTp8X6mJgYeHp6AgCcnZ3FS2Ofdq9nQkIC+vbtC11dXRgYGGDIkCG4fPmyWD9+/Hj069cPAODp6QmJRAInJ6enxrtv3z4UFhbC09MTo0aNwt69e/Hw4UOldhKJBNOmTcN3332HDh06QCqVws7OTulS7Pz8fMyYMQMWFhaQSqVo1qwZ3n77bZw7dw4AEBERAXV1dYXLm1etWgWJRIKgoCCxrKysDDKZDMHBwWJZeXk5Pv/8c9jZ2aFRo0Zo3rw5PvjgA2RnZyvEUHGP8eHDh9GtWzdoa2tj06ZNVc7/v//+Q15eHvr06VNlfbNmzcTfq7rHu+J4SU9Ph5ubGxo3bowWLVpg3bp1AIALFy7grbfegq6uLlq3bo0dO3YojF/Te31XrlyJ3r17w9jYGNra2nBwcMDu3buV2lW8T9u3b4ednR2kUqn4HlW+xzs0NBSzZ88GALRp00Y8DtPS0tCvXz907ty5yjhsbGzg6ur6xFgrxMfHo0uXLmjUqBHat2+PvXv3inXXr1+HRCJBeHi4Ur+TJ09CIpHgm2++qdF2niQhIQFqamr45JNPFMp37NgBiUSCDRs2iGWV952NjQ0aNWoEBwcHJCUlKY17+/Zt+Pj4oHnz5uLfQlRUlEKbivc2NjYWH3/8MVq0aAEdHR3k5eVV+b47OTmhQ4cO+OOPP9CvXz/o6OjAyspKfJ9/+ukn9OzZE9ra2rCxscGRI0eeK65du3Zh6dKlaNmyJRo1aoT+/fsjJSVFIZ64uDjcvHlTPD5UsfYAvQQEIiIiapCio6MFAMKRI0eEf//9V7h7965w8eJF4YMPPhDU1NSE+Ph4sW15ebnw1ltvCRKJRJg4caKwdu1awd3dXQAgzJgxQ2x3+vRpQV1dXQgMDBTLRo0aJWhrawt//fWXWDZu3DihUaNGgrW1tTBmzBhh7dq1gpubmwBAmD9/vkKcAIQFCxaIry9evCjo6uoKpqamwuLFi4XPPvtMaNOmjSCVSoXTp08LgiAIqampQkBAgABAmDdvnrB161Zh69atQmZmZrX748cffxQ0NDQEuVwuLF++XFi4cKHQpEkTwdDQULhx44YgCIJw8uRJYd68eQIAISAgQNi6davCfqrOgAEDhP79+wuCIAg3b94UJBKJsGvXLqV2AITOnTuLc/v8888FS0tLQUdHR/jvv//EdqNHjxa0tLSEoKAg4auvvhKWLVsmuLu7C9u2bRMEQRDOnTsnABD+97//iX2GDBkiqKmpCd26dRPLzpw5IwAQ9u/fL5ZNnDhR0NDQECZNmiRs3LhRCA4OFnR1dYXu3bsLxcXFYrvWrVsLVlZWgqGhoTB37lxh48aNQmJiYpXzLysrE7S1tQUHBwchKyvrifvqxo0bAgAhOjpaLKs4Xtq3by9MmTJFWLdundC7d2+xnZmZmTB79mzhiy++EOzs7AR1dXXh+vXrYv/ExEQBgEJ848aNE1q3bq2w7ZYtWwp+fn7C2rVrhdWrVws9evRQ2j+C8Oh9srW1FZo2bSosXLhQWLdunfD777+LdRXH6/nz5wUvLy8BgBAeHi4ehwUFBUJkZKQAQLhw4YLC2L/++qsAQNiyZcsT91Pr1q0FuVwuGBgYCHPnzhVWr14tdOzYUelvt0+fPoKDg4NSfz8/P0Emkwn379+vdhsV78XChQuFf//9V+mnpKREbOvv7y9oaGgIv/32myAIgnDnzh3ByMhIcHFxEcrLyxX2XYcOHYQmTZoIixYtEpYtWya0bt1a0NbWVtgXmZmZQsuWLQVzc3Nh0aJFwoYNG4TBgweL+7JCxXvbvn17oUuXLsLq1auFsLAw4f79+1W+7/369RPMzMwEc3Nz8Zhp3769oK6uLsTGxgomJiZCaGio8PnnnwstWrQQ9PX1hby8vGeOy97eXnBwcBDCw8OF0NBQQUdHR+jRo4fYLj4+XujSpYvQpEkT8fjYt29fte8JvbqYeBMRETVQFYn34z9SqVSIiYlRaPvdd98JAIQlS5YolI8YMUKQSCRCSkqKWBYSEiKoqakJSUlJwrfffisAED7//HOFfuPGjRMACNOnTxfLysvLhXfffVfQ0tIS/v33X7H88cR76NChgpaWlpCamiqW3blzR5DJZIKjo6NYVrHt6pLBx3Xp0kVo1qyZQmJ4/vx5QU1NTRg7dqxYVvGB+ttvv63RuP/884+goaEhREZGimW9e/cWhgwZotQWgKClpaWwP8+fPy8AEL744guxTF9fX/D39692m2VlZYKenp4wZ84cQRAe7VtjY2PB09NTUFdXF/Lz8wVBEITVq1cLampqQnZ2tiAIgnD8+HEBgLB9+3aF8Q4dOqRU3rp1awGAcOjQoRrth08++UQAIOjq6goDBw4Uli5dKiZplVWXeAMQPv30U7EsOztb0NbWFiQSiRAbGyuWX7lyRemYqWni/eDBA4XXxcXFQocOHYS33npLoRyAoKamJvz5559K8T++7RUrVggAxC9vKuTk5AiNGjUSgoODFcoDAgIEXV1doaCgQGnsyir2/549e8Sy3NxcwdTUVLC3txfLNm3aJAAQLl++rDCvJk2aCOPGjXviNirei+p+Tp06Jba9f/++YGVlJdjZ2QkPHz4U3n33XUFPT0+4efOm0v4BIJw9e1Ysu3nzptCoUSNh2LBhYpmvr69gamqq8IWTIDz6Ik9fX198ryreW0tLS6X3r7rEG4CwY8cOsazimFFTUxO/vBMEQTh8+LDSsVjbuGxtbYWioiKx3Zo1a5S+cHn33XeVjkV6/fBScyIiogZu3bp1+PHHH/Hjjz9i27ZtcHZ2xsSJExUuWT1w4ADU1dUREBCg0HfmzJkQBEFhFfTQ0FDY2dlh3Lhx8PPzQ79+/ZT6VZg2bZr4e8UlqMXFxVVe3gk8ujQ6Pj4eQ4cOhaWlpVhuamqK0aNH4+eff0ZeXl6t90FGRgaSk5Mxfvx4GBkZieWdOnXC22+/jQMHDtR6zAqxsbFQU1ODh4eHWObl5YWDBw8qXb4NAC4uLmjbtq1CDHp6erh+/bpYZmBggF9++QV37typcptqamro3bu3ePnu5cuXkZWVhblz50IQBJw6dQoAcPz4cXTo0EG8X/3bb7+Fvr4+3n77bfz333/ij4ODAxo3bozExESF7bRp06bGl0QvXLgQO3bsgL29PQ4fPoyPPvoIDg4O6Nq1q8Ll/E9SeTE7AwMD2NjYQFdXFyNHjhTLbWxsYGBgoLC/aqry/enZ2dnIzc1F3759xUv4K+vXrx/at29f621U0NfXx5AhQ/DNN9+It1eUlZVh586dGDp0KHR1dZ86hpmZGYYNGya+1tPTw9ixY/H7778jMzMTADBy5Eg0atQI27dvF9sdPnwY//33X43v2548ebL4b0Tln8rz19HRQUxMDC5fvgxHR0fExcUhPDwcrVq1UhrvjTfegIODg/i6VatWGDJkCA4fPoyysjIIgoA9e/bA3d0dgiAoHIuurq7Izc1Vek/GjRtX5foCVWncuDFGjRolvq44ZmxtbdGzZ0+xvOL3imPpWeKaMGGCwuJuffv2VRiTGg4m3kRERA1cjx494OLiAhcXF3h7eyMuLg7t27cXk2AAuHnzJszMzCCTyRT62traivUVtLS0EBUVhRs3biA/Px/R0dGQSCRK21VTU1NIngFALpcDQLWPAPv333/x4MED2NjYKNXZ2tqivLwcf//9d80n//9VxF/duP/99x/u379f63EBYNu2bejRoweysrKQkpKClJQU2Nvbo7i4GN9++61S+6oSFUNDQ4Ukffny5bh48SLMzc3Ro0cPhIaGKn2Q79u3L3777TcUFhbi+PHjMDU1RdeuXdG5c2ccP34cAPDzzz+LiQAAXLt2Dbm5uWjWrBmaNm2q8FNQUKCwCBrwKPGuDS8vLxw/fhzZ2dmIj4/H6NGj8fvvv8Pd3b3Ke94ra9SoEZo2bapQpq+vj5YtWyodX/r6+lV+qfE0+/fvR69evdCoUSMYGRmhadOm2LBhA3Jzc5Xa1nbuVRk7dizS09PF9+PIkSP4559/MGbMmBr1t7KyUpr7439DBgYGcHd3V7jvffv27WjRogXeeuutGm3H2tpa/Dei8o+enp5Cuz59+mDq1Kn49ddf4erqCh8fn2rHe5xcLseDBw/w77//4t9//0VOTg6+/PJLpeNwwoQJAPBcx2J1x4y5ublSGQDxWHqWuB7/ezY0NFQYkxoOrmpORERECtTU1ODs7Iw1a9bg2rVrsLOzq/UYhw8fBvBodeFr1669kCTlVXTt2jWcOXMGQNXJxvbt2zF58mSFMnV19SrHqjgrCjw6i9m3b1/s27cP8fHxWLFiBZYtW4a9e/di4MCBAIA333wTJSUlOHXqFI4fPy4m2H379sXx48dx5coV/PvvvwqJd3l5OZo1a6ZwdrSyxxPfmp5hfJyenh7efvttvP3229DU1MTmzZvxyy+/iAvXVaW6/VKT/VUTx48fx+DBg+Ho6Ij169fD1NQUmpqaiI6OVlqsDXj2uVfm6uqK5s2bY9u2bXB0dMS2bdtgYmLywle8Hzt2LL799lucPHkSHTt2xA8//AA/Pz+oqb3Yc3BFRUXiQmapqal48OABdHR0aj1OxXPu33//fYwbN67KNp06dVJ4XZv341mPpWeJ60Udn/TqY+JNRERESkpLSwEABQUFAB49W/nIkSPIz89XOOt95coVsb7CH3/8gUWLFmHChAlITk7GxIkTceHCBfHsUYXy8nJcv35dPEMHAFevXgWAalf1bdq0KXR0dPDXX38p1V25cgVqamriWauqzrJXpyL+6sZt0qRJjS79fdz27duhqamJrVu3Kn0A//nnnxEREYH09PQqz3I/jampKfz8/ODn54e7d++ia9euWLp0qZh49+jRA1paWjh+/DiOHz8urq7t6OiIyMhIHD16VHxdoW3btjhy5Aj69OnzQhLLmujWrRs2b96MjIyMOtledfbs2YNGjRrh8OHDkEqlYnl0dPRzjfuk41BdXR2jR49GTEwMli1bhu+++w6TJk2qNll7XEpKCgRBUNhGVX9DAwYMQNOmTbF9+3b07NkTDx48qPFZ9dpYsGABLl++jJUrVyI4OBhz585FRESEUrtr164plV29ehU6OjrilzsymQxlZWUv1WP3mjZtqpK4avNvFb26eKk5ERERKSgpKUF8fDy0tLTES8kHDRqEsrIyrF27VqFteHg4JBKJmOyVlJRg/PjxMDMzw5o1axATE4N//vkHgYGBVW6r8niCIGDt2rXQ1NRE//79q2yvrq6Od955B99//73C5ej//PMPduzYgTfffFO8/LUiUa78SK3qmJqaokuXLti8ebNC+4sXLyI+Ph6DBg166hhV2b59O/r27Yv33nsPI0aMUPipSIRr+zinsrIypUufmzVrBjMzMxQVFYlljRo1Qvfu3fHNN98gPT1d4Yx3YWEhIiIi0LZtW5iamop9Ro4cibKyMixevFhpu6WlpTXal1V58OCBeF/54yrWB6jqMv+6pK6uDolEIj7yDnh0ufZ33333XOM+7TgcM2YMsrOz8cEHH6CgoKBWz8u+c+cO9u3bJ77Oy8vDli1b0KVLF5iYmIjlGhoa8PLywq5duxATE4OOHTsqnZl9Xr/88gtWrlyJGTNmYObMmZg9ezbWrl2Ln376SantqVOnFO6F/vvvv/H999/jnXfegbq6OtTV1eHh4YE9e/bg4sWLSv3//fffFxp7TakqLl1d3SpvZ6DXC894ExERNXAHDx4Uz1zfvXsXO3bswLVr1zB37lwxiXV3d4ezszM++ugjpKWloXPnzoiPj8f333+PGTNmiIuBLVmyBMnJyTh69ChkMhk6deqETz75BB9//DFGjBihkMA2atQIhw4dwrhx49CzZ08cPHgQcXFxmDdvntIlzZUtWbIEP/74I9588034+flBQ0MDmzZtQlFREZYvXy6269KlC9TV1bFs2TLk5uZCKpXirbfeUnhmdGUrVqzAwIED8cYbb8DX1xeFhYX44osvoK+vLz6XuTZ++eUXpKSkKCwgV1mLFi3QtWtXbN++XeE52k+Tn5+Pli1bYsSIEejcuTMaN26MI0eO4MyZM1i1apVC2759++Kzzz6Dvr4+OnbsCOBRkm5jY4O//voL48ePV2jfr18/fPDBBwgLC0NycjLeeecdaGpq4tq1a/j222+xZs0ajBgxonY7Ao8S7969e6NXr14YMGAAzM3NkZOTg++++w7Hjx/H0KFDYW9vX+txX6R3330Xq1evxoABAzB69GjcvXsX69atg5WVFf74449nHrdiEbGPPvoIo0aNgqamJtzd3cWE3N7eHh06dMC3334LW1tbdO3atcZjy+Vy+Pr64syZM2jevDmioqLwzz//VHmWfuzYsYiIiEBiYiKWLVtWqzmcO3cO27ZtUypv27Yt3njjDTx8+BDjxo2DtbU1li5dCuDRYnr/+9//MGHCBFy4cEHhipEOHTrA1dUVAQEBkEqlWL9+vdinwmeffYbExET07NkTkyZNQvv27XHv3j2cO3cOR44cwb1792o1hxdFFXE5ODhg586dCAoKQvfu3dG4cWO4u7urIHqqV/WyljoRERHVu6oeJ9aoUSOhS5cuwoYNGxSevSsIgpCfny8EBgYKZmZmgqampmBtbS2sWLFCbPfbb78JGhoaCo8IEwRBKC0tFbp37y6YmZmJj60aN26coKurK6SmpgrvvPOOoKOjIzRv3lxYsGCBUFZWptAfjz2eSRAePafa1dVVaNy4saCjoyM4OzsLJ0+eVJpjZGSkYGlpKairq9fo0WJHjhwR+vTpI2hrawt6enqCu7u7cOnSJYU2NX2c2PTp0wUACo89e1xoaKgAQDh//rw416oeE9a6dWvx0U9FRUXC7Nmzhc6dOwsymUzQ1dUVOnfuLKxfv16pX1xcnABAGDhwoEL5xIkTBQDC119/XWVcX375peDg4CBoa2sLMplM6NixozBnzhzhzp07CjG9++67T9wHFUpKSoTIyEhh6NChQuvWrQWpVCro6OgI9vb2wooVKxQet1Td48R0dXWVxu3Xr59gZ2enVP54bDV9nNjXX38tWFtbC1KpVGjXrp0QHR0tLFiwQHj8I3N171NF3ePH6+LFi4UWLVoIampqVT5abPny5UqPS3uaijkePnxY6NSpkxjzk45LOzs7QU1NTbh161aNtvG0x4lVHJOBgYGCurq68Msvvyj0P3v2rKChoSFMnTpVLKvYd9u2bRP3tb29fZV/m//884/g7+8vmJubC5qamoKJiYnQv39/4csvvxTbPOnvsbrHidXkmHk83hcVV1XHd0FBgTB69GjBwMBAAMBHi72mJILAO/uJiIiobo0fPx67d+8W7yEnasjWrFmDwMBApKWlPdP9/jVlb28PIyMj8f7++iCRSODv76902wrR6473eBMRERER1RNBEPD111+jX79+Kk26z549i+TkZIwdO1Zl2yCi6vEebyIiIiKiOnb//n388MMPSExMxIULF/D999+rZDsXL17Eb7/9hlWrVsHU1BTvvfeeSrZDRE/GxJuIiIiIqI79+++/GD16NAwMDDBv3jwMHjxYJdvZvXs3Fi1aBBsbG3zzzTdo1KiRSrZDRE/Ge7yJiIiIiIiIVIj3eBMRERERERGpEBNvIiIiIiIiIhXiPd5ERPTClJeX486dO5DJZJBIJPUdDhEREZHKCIKA/Px8mJmZQU3tyee0mXgTEdELc+fOHZibm9d3GERERER15u+//0bLli2f2IaJNxERvTAymQzAo/8B6enp1XM0RERERKqTl5cHc3Nz8fPPkzDxJiKiF6bi8nI9PT0m3kRERNQg1OT2Oi6uRkRERERERKRCPONNRERERPQay8rKQn5+fn2HoVIymQzGxsb1HQZRtZh4ExERERG9prKysjA3JAQlxcX1HYpKaWpp4bOwMCbf9NJi4k1ERERE9JrKz89HSXEx2rr0g7aRfr3EUJidg9Qfk9D2bUdoGxq8+PHv5SL1yE/Iz89n4k0vLSbeRERERESvOW0jfeg2bVK/MRga1HsMRPWFi6sRERERERERqRATbyIiIiIiIiIVYuJNREREREREpEJMvImIiIiIiIhUiIk3ERERERERkQox8SYiIiIiek5FRUVIS0tDUVFRfYdCKsb3mp4FE28iIiIioueUkZGB0NBQZGRk1HcopGJ8r+lZMPEmIiIiIiIiUiEm3kREREREREQqxMSb6CUlkUjw3XffqXw7x44dg0QiQU5Ojsq31VCkpaVBIpEgOTm5vkMhIiIiopcAE2+iepCZmYnp06fD0tISUqkU5ubmcHd3x9GjR+s8lt69eyMjIwP6+vq17rtnzx6oq6vj9u3bVdZbW1sjKCgIAODk5ASJRAKJRAKpVIoWLVrA3d0de/fuVepX0U4ikUBDQwOtWrVCUFCQyhcxkUgkaNSoEW7evKlQPnToUIwfP77G45ibmyMjIwMdOnR4wREqqryf9PT00L17d3z//fdVtg0LC4O6ujpWrFhRZX1mZiY+/PBDWFlZoVGjRmjevDn69OmDDRs24MGDB6qcBhEREdFrj4k3UR1LS0uDg4MDEhISsGLFCly4cAGHDh2Cs7Mz/P396zweLS0tmJiYQCKR1Lrv4MGDYWxsjM2bNyvVJSUlISUlBb6+vmLZpEmTkJGRgdTUVOzZswft27fHqFGjMHnyZKX+0dHRyMjIwI0bN7B+/Xps3boVS5YsqXFsFWeda0sikeCTTz6pdb/K1NXVYWJiAg0NjecapyYq9tPZs2fRp08fjBgxAhcuXFBqFxUVhTlz5iAqKkqp7vr167C3t0d8fDw+/fRT/P777zh16hTmzJmD/fv348iRIyqfBxEREdHrjIk3UR3z8/ODRCLBr7/+Cg8PD8jlctjZ2SEoKAinT5+utl9wcDDkcjl0dHRgaWmJ+fPno6SkRKw/f/48nJ2dIZPJoKenBwcHB5w9exYAcPPmTbi7u8PQ0BC6urqws7PDgQMHAFR9qfmJEyfg5OQEHR0dGBoawtXVFdnZ2UoxaWpqYsyYMYiJiVGqi4qKQs+ePWFnZyeW6ejowMTEBC1btkSvXr2wbNkybNq0CZGRkUrJnYGBAUxMTGBubg43NzcMGTIE586dq9E+fh7Tpk3Dtm3bcPHixWrbHDp0CG+++SYMDAxgbGwMNzc3pKamivWVLzUvLy9Hy5YtsWHDBoUxfv/9d6ipqYln13NycjBx4kQ0bdoUenp6eOutt3D+/Pmnxluxn+RyORYvXozS0lIkJiYqtPnpp59QWFiIRYsWIS8vDydPnlSo9/Pzg4aGBs6ePYuRI0fC1tYWlpaWGDJkCOLi4uDu7v7UOIiIiIioeqo/HUNEonv37uHQoUNYunQpdHV1leoNDAyq7SuTyRATEwMzMzNcuHABkyZNgkwmw5w5cwAA3t7esLe3x4YNG6Curo7k5GRoamoCAPz9/VFcXIykpCTo6uri0qVLaNy4cZXbSU5ORv/+/eHj44M1a9ZAQ0MDiYmJKCsrq7K9r68vVq9ejaSkJDg6OgIACgoKsHv3boSHhz91n4wbNw4zZ87E3r174eLiUmWbq1evIiEhoVaXez+rPn364OrVq5g7dy72799fZZv79+8jKCgInTp1QkFBAT755BMMGzYMycnJUFNT/D5TTU0NXl5e2LFjB6ZOnSqWb9++HX369EHr1q0BAJ6entDW1sbBgwehr6+PTZs2oX///rh69SqMjIyeGndpaSm+/vprAI+uYqjs66+/hpeXFzQ1NeHl5YWvv/4avXv3BgBkZWWJZ7qrOiYBPPHKgaKiIoVbAPLy8p4aKxHR6+zOnTv1HYKCly0eVaqruTakfUovDhNvojqUkpICQRDQrl27Wvf9+OOPxd8tLCwwa9YsxMbGiol3eno6Zs+eLY5tbW0ttk9PT4eHhwc6duwIALC0tKx2O8uXL0e3bt2wfv16sazyWevHtW/fHr169UJUVJSYeO/atQuCIGDUqFFPnZeamhrkcjnS0tIUyr28vKCuro7S0lIUFRXBzc0NISEhTx3vRQgLC0OnTp1w/Phx9O3bV6new8ND4XVUVBSaNm2KS5cuVXlft7e3N1atWoX09HS0atUK5eXliI2NFd/Tn3/+Gb/++ivu3r0LqVQKAFi5ciW+++477N69u8pL8StU7KfCwkKUl5fDwsICI0eOFOvz8vKwe/dunDp1CgDw/vvvo2/fvlizZg0aN24sHpM2NjYK4zZp0gQPHz4E8OiLm2XLllW7rxYuXFhtfEREDc2XX35Z3yE0WNz39DJj4k1UhwRBeOa+O3fuREREBFJTU1FQUIDS0lLo6emJ9UFBQZg4cSK2bt0KFxcXeHp6om3btgCAgIAATJ06FfHx8XBxcYGHhwc6depU5XaSk5Ph6elZq9h8fHwQGBiIL774AjKZDFFRUfD09IRMJqtRf0EQlM6qhoeHw8XFBWVlZUhJSUFQUBDGjBmD2NjYasexs7MTL92u2NeVz+z37dsXBw8efGo87du3x9ixYzF37lycOHFCqf7atWv45JNP8Msvv+C///5DeXk5gEdfcFSVeHfp0gW2trbYsWMH5s6di59++gl3794V9/P58+dRUFAAY2NjhX6FhYUKl7BXpWI/Xb9+HYGBgYiIiFA4Q/7NN9+gbdu26Ny5sxhL69atsXPnToX77x/366+/ory8HN7e3k9c1C4kJERcQA94lOibm5s/MWYiotfZ5MmTYWZmVt9hiO7cudNgEtK62vcNaZ/Si8PEm6gOWVtbQyKR4MqVK7Xqd+rUKXh7e2PhwoVwdXWFvr4+YmNjsWrVKrFNaGgoRo8ejbi4OBw8eBALFixAbGwshg0bhokTJ8LV1RVxcXGIj49HWFgYVq1ahenTpyttS1tbu9bzGjVqFAIDA7Fr1y44OjrixIkTCAsLq1HfsrIyXLt2Dd27d1coNzExgZWVFQDAxsYG+fn58PLywpIlS8Tyxx04cEC87/327dtwcnJSeKRXbea2cOFCyOXyKh/p5u7ujtatWyMyMhJmZmYoLy9Hhw4dUFxcXO143t7eYuK9Y8cODBgwQEy0CwoKYGpqimPHjin1e9LtB8D/7ScrKytER0dj0KBBuHTpEpo1awbg0WXmf/75p8JCb+Xl5YiKioKvry+srKwgkUjw119/KYxbcVXE0/aZVCoVz9ITERFgZmYGCwuL+g6jQeK+p5cZF1cjqkNGRkZwdXXFunXrcP/+faX66p6lffLkSbRu3RofffQRunXrBmtra6VHXgGAXC5HYGAg4uPjMXz4cERHR4t15ubmmDJlCvbu3YuZM2ciMjKyym116tSp1o81k8lk8PT0RFRUFKKjoyGXy6u8RLsqmzdvRnZ2ttLl249TV1cH8OgscHVat24tJqEV905XvLayskKLFi1qOKNH+2vatGmYN2+ewv3tWVlZ+Ouvv/Dxxx+jf//+sLW1rXLhuceNHj0aFy9exG+//Ybdu3fD29tbrOvatSsyMzOhoaGhEK+VlRWaNGlS45h79OgBBwcHLF26FABw4cIFnD17FseOHUNycrL4c+zYMZw6dQpXrlyBsbEx3n77baxdu7bKY5KIiIiInh8Tb6I6tm7dOpSVlaFHjx7Ys2cPrl27hsuXLyMiIgJvvPFGlX2sra2Rnp6O2NhYpKamIiIiAvv27RPrCwsLMW3aNBw7dgw3b97EiRMncObMGdja2gIAZsyYgcOHD+PGjRs4d+4cEhMTxbrHhYSE4MyZM/Dz88Mff/yBK1euYMOGDfjvv/+eOC9fX1+cPHkSGzduhI+PT5VtHjx4gMzMTNy6dQunT59GcHAwpkyZgqlTp8LZ2VmhbU5ODjIzM3Hnzh389NNPWLRoEeRyebVxq0JISAju3LmjsOK6oaEhjI2N8eWXXyIlJQUJCQkKl1pXx8LCAr1794avry/KysowePBgsc7FxQVvvPEGhg4divj4eKSlpeHkyZP46KOPxJXpa2rGjBnYtGkTbt++ja+//ho9evSAo6MjOnToIP44Ojqie/fu4mJs69evR2lpKbp164adO3fi8uXL+Ouvv7Bt2zZcuXJF/NKDiIiIiJ4NE2+iOmZpaYlz587B2dkZM2fORIcOHfD222/j6NGjSo+cqjB48GAEBgZi2rRp6NKlC06ePIn58+eL9erq6sjKysLYsWMhl8sxcuRIDBw4UFz0qqysDP7+/rC1tcWAAQMgl8sVFk+rTC6XIz4+HufPn0ePHj3wxhtv4Pvvv3/qM6nffPNN2NjYIC8vD2PHjq2yTWRkJExNTdG2bVsMHz4cly5dws6dO6uMZcKECTA1NUXLli3h5eUFOzs7HDx4sE6ejV3ByMgIwcHB4iJjwKPF4GJjY/Hbb7+hQ4cOCAwMxIoVK2o0nre3N86fP49hw4YpXMItkUhw4MABODo6YsKECZDL5Rg1ahRu3ryJ5s2b1yrmAQMGoE2bNli6dCm2bdtW7ZUEHh4e2LJlC0pKStC2bVv8/vvvcHFxQUhICDp37oxu3brhiy++wKxZs7B48eJaxUBEREREiiTC86z2REREVEleXh709fWRm5ursPgfEdHrLi0tDaGhoQgNDX2p7jOuiKvDyMHQbVrz25depPv//oeLu35QWQwV49fVvn9Z32uqe7X53MMz3kREREREREQqxMSbiIiIiIiISIWYeBMRERERERGpEBNvIiIiIqLnZGpqitDQUJiamtZ3KKRifK/pWdTd8sBERERERK8pqVTKhbYaCL7X9Cx4xpuIiIiIiIhIhZh4ExEREREREakQE28iIiIiIiIiFWLiTURERERERKRCTLyJiIiIiIiIVIirmhMRERERveYK7+XW37azcxT++8LHr8e5EdUUE28iIiIioteUTCaDppYWUo/8VN+hIPXHJJWNramlBZlMprLxiZ6XRBAEob6DICKi10NeXh709fWRm5sLPT29+g6HiIgAZGVlIT8/v77DUCmZTAZjY+P6DoMamNp87uEZbyIiIiKi15ixsTGTUqJ6xsXViIiIiIiIiFSIiTcRERERERGRCjHxJiIiIiIiIlIhJt5EREREREREKsTEm4iIiIiIiEiFuKo5ERERERG99F6lx6Lx8Wb0OCbeRERERET0UsvKykJISAiKi4vrO5Qa0dLSQlhYGJNvEjHxJiIiIiKil1p+fj6Ki4sxYUIfmJjoP7V9ZmYuoqNP1Lj9i1Sx7fz8fCbeJGLiTURERERErwQTE320alXzZLa27YlUhYurEREREREREakQE28iIiIiIiIiFWLiTURERERERKRCTLyJiIiIiIiIVIiJNxEREREREZEKMfEmIiIiInrFFBUVIS0tDUVFRfUdCr0keEy83Jh4ExERERG9YjIyMhAaGoqMjIz6DoVeEjwmXm5MvImIiIiIiIhUiIk3ERERERERkQox8SaqAYlEgu+++07l2zl27BgkEglycnJUvq2GwsLCAp9//nl9h/HKGD9+PIYOHVrfYRARERG9Vph4U4OXmZmJ6dOnw9LSElKpFObm5nB3d8fRo0frPJbevXsjIyMD+vr6te67Z88eqKur4/bt21XWW1tbIygoCADg5OQEiUQCiUQCqVSKFi1awN3dHXv37lXqV9FOIpFAQ0MDrVq1QlBQUJ0s3JGYmIhBgwbB2NgYOjo6aN++PWbOnFntHF92aWlpCvvTyMgI/fr1w/Hjx+s7NCIiIiJSISbe1KClpaXBwcEBCQkJWLFiBS5cuIBDhw7B2dkZ/v7+dR6PlpYWTExMIJFIat138ODBMDY2xubNm5XqkpKSkJKSAl9fX7Fs0qRJyMjIQGpqKvbs2YP27dtj1KhRmDx5slL/6OhoZGRk4MaNG1i/fj22bt2KJUuW1Di2ioSzNjZt2gQXFxeYmJhgz549uHTpEjZu3Ijc3FysWrWqVmO9aMXFxc/V/8iRI8jIyEBSUhLMzMzg5uaGf/75p97iISIiIiLVYuJNDZqfnx8kEgl+/fVXeHh4QC6Xw87ODkFBQTh9+nS1/YKDgyGXy6GjowNLS0vMnz8fJSUlYv358+fh7OwMmUwGPT09ODg44OzZswCAmzdvwt3dHYaGhtDV1YWdnR0OHDgAoOpLzU+cOAEnJyfo6OjA0NAQrq6uyM7OVopJU1MTY8aMQUxMjFJdVFQUevbsCTs7O7FMR0cHJiYmaNmyJXr16oVly5Zh06ZNiIyMxJEjRxT6GxgYwMTEBObm5nBzc8OQIUNw7ty5Gu3jZ3Hr1i0EBAQgICAAUVFRcHJygoWFBRwdHfHVV1/hk08+Edvu2bMHdnZ2kEqlsLCwqDIpf/DgAXx8fCCTydCqVSt8+eWXCvV///03Ro4cCQMDAxgZGWHIkCFIS0sT6ysuv166dCnMzMxgY2NTo37VMTY2homJCTp06IB58+YhLy8Pv/zyi1h/8eJFDBw4EI0bN0bz5s0xZswY/Pfff2K9k5MTpk2bhhkzZqBJkyZwdXUFAKxevRodO3aErq4uzM3N4efnh4KCArFfTEwMDAwMcPjwYdja2qJx48YYMGDAE1c/PXPmDJo2bYply5Y9dV5EREREVDWN+g6AqL7cu3cPhw4dwtKlS6Grq6tUb2BgUG1fmUyGmJgYmJmZ4cKFC5g0aRJkMhnmzJkDAPD29oa9vT02bNgAdXV1JCcnQ1NTEwDg7++P4uJiJCUlQVdXF5cuXULjxo2r3E5ycjL69+8PHx8frFmzBhoaGkhMTERZWVmV7X19fbF69WokJSXB0dERAFBQUIDdu3cjPDz8qftk3LhxmDlzJvbu3QsXF5cq21y9ehUJCQkYP378U8d7Vt9++y2Ki4vF/fm4ivfmt99+w8iRIxEaGor33nsPJ0+ehJ+fH4yNjRXiW7VqFRYvXox58+Zh9+7dmDp1Kvr16wcbGxuUlJTA1dUVb7zxBo4fPw4NDQ0sWbIEAwYMwB9//AEtLS0AwNGjR6Gnp4cff/wRAGrc70kKCwuxZcsWABDb5+Tk4K233sLEiRMRHh6OwsJCBAcHY+TIkUhISBD7bt68GVOnTsWJEyfEMjU1NURERKBNmza4fv06/Pz8MGfOHKxfv15s8+DBA6xcuRJbt26Fmpoa3n//fcyaNQvbt29Xii8hIQHDhw/H8uXLq7wSAnj0zNDKtx3k5eU9dd5ERPTi3Llzp75DqBOv4jzrOuZXcR81JEy8qcFKSUmBIAho165drft+/PHH4u8WFhaYNWsWYmNjxUQxPT0ds2fPFse2trYW26enp8PDwwMdO3YEAFhaWla7neXLl6Nbt24KiVPls9aPa9++PXr16oWoqCgx8d61axcEQcCoUaOeOi81NTXI5XKls7ZeXl5QV1dHaWkpioqK4ObmhpCQkKeO96yuXbsGPT09mJqaPrHd6tWr0b9/f8yfPx8AIJfLcenSJaxYsUIh8R40aBD8/PwAPLpaITw8HImJibCxscHOnTtRXl6Or776SrwcPjo6GgYGBjh27BjeeecdAICuri6++uorMUHetm1bjfpVpXfv3lBTU8ODBw8gCAIcHBzQv39/AMDatWthb2+PTz/9VGwfFRUFc3NzXL16FXK5HMCjY2r58uUK486YMUP83cLCAkuWLMGUKVMUjp+SkhJs3LgRbdu2BQBMmzYNixYtUopx3759GDt2LL766iu899571c4lLCwMCxcurLaeiIhU6/GruOjlwfeGKmPiTQ2WIAjP3Hfnzp2IiIhAamoqCgoKUFpaCj09PbE+KCgIEydOxNatW+Hi4gJPT08x0QkICMDUqVMRHx8PFxcXeHh4oFOnTlVuJzk5GZ6enrWKzcfHB4GBgfjiiy8gk8kQFRUFT09PyGSyGvUXBEHpfuzw8HC4uLigrKwMKSkpCAoKwpgxYxAbG1vtOHZ2drh586Y4JgCFM/t9+/bFwYMHaxxDVS5fvowhQ4YolPXp0weff/45ysrKoK6uDgAK+1cikcDExAR3794F8Oi2gJSUFKX98/DhQ6SmpoqvO3bsqHAWu6b9qrJz5060a9cOFy9exJw5cxATEyNeEXH+/HkkJiZWeRVEamqqmHg7ODgo1R85cgRhYWG4cuUK8vLyUFpaiocPH+LBgwfQ0dEB8OgWg4pjEQBMTU3FfVHhl19+wf79+7F79+6nrnAeEhIiLtoHPDrjbW5u/sQ+RET04kyePBlmZmb1HYbK3blz55VLZOv6vXkV91FDwsSbGixra2tIJBJcuXKlVv1OnToFb29vLFy4EK6urtDX10dsbKzCvcWhoaEYPXo04uLicPDgQSxYsACxsbEYNmwYJk6cCFdXV8TFxSE+Ph5hYWFYtWoVpk+frrQtbW3tWs9r1KhRCAwMxK5du+Do6IgTJ04gLCysRn3Lyspw7do1dO/eXaHcxMQEVlZWAAAbGxvk5+fDy8sLS5YsEcsfd+DAAfG+99u3b8PJyQnJyck1mptcLkdubi4yMjKeeta7JiqS2goSiQTl5eUAHl2K7+DgUOWl1k2bNhV/f/x2hJr2q4q5uTmsra1hbW2N0tJSDBs2DBcvXoRUKkVBQQHc3d2rvKe68r54PJ60tDS4ublh6tSpWLp0KYyMjPDzzz/D19cXxcXFYuJd1b54/Euotm3bwtjYGFFRUXj33XeV+lQmlUohlUqfOF8iIlIdMzMzWFhY1HcYVAW+N1QZF1ejBsvIyAiurq5Yt24d7t+/r1Rf3bO0T548idatW+Ojjz5Ct27dYG1tLZ7ZrUwulyMwMBDx8fEYPnw4oqOjxTpzc3NMmTIFe/fuxcyZMxEZGVnltjp16lTrx5rJZDJ4enoiKioK0dHRkMvl6Nu3b436bt68GdnZ2fDw8Hhiu4ozyYWFhdW2ad26NaysrGBlZYXWrVsDgPjaysoKLVq0qLbviBEjoKWlpXQpdYWK98bW1lbhHmfg0WJ0crlcjPFpunbtimvXrqFZs2YK8VlZWT3xsW7P2u9xI0aMgIaGhng5eNeuXfHnn3/CwsJCadyq1iKo8Ntvv6G8vByrVq1Cr169IJfLn/leryZNmiAhIQEpKSkYOXKkwsKBRERERFR7TLypQVu3bh3KysrQo0cP7NmzB9euXcPly5cRERGBN954o8o+1tbWSE9PR2xsLFJTUxEREYF9+/aJ9YWFhZg2bRqOHTuGmzdv4sSJEzhz5gxsbW0BPLoP9/Dhw7hx4wbOnTuHxMREse5xISEhOHPmDPz8/PDHH3/gypUr2LBhg8IK11Xx9fXFyZMnsXHjRvj4+FTZ5sGDB8jMzMStW7dw+vRpBAcHY8qUKZg6dSqcnZ0V2ubk5CAzMxN37tzBTz/9hEWLFkEul1cb9/MyNzdHeHg41qxZA19fX/z000/ivvzggw+wePFiAMDMmTNx9OhRLF68GFevXsXmzZuxdu1azJo1q8bb8vb2RpMmTTBkyBAcP34cN27cwLFjxxAQEIBbt2698H6Pk0gkCAgIwGeffYYHDx7A398f9+7dg5eXF86cOYPU1FQcPnwYEyZMqHZRPeDRlxolJSX44osvcP36dWzduhUbN26scRyPa9asGRISEnDlyhV4eXmhtLT0mcciIiIiauiYeFODZmlpiXPnzsHZ2RkzZ85Ehw4d8Pbbb+Po0aPYsGFDlX0GDx6MwMBATJs2DV26dMHJkyfFxb2AR2eDs7KyMHbsWMjlcowcORIDBw4UF6AqKyuDv78/bG1tMWDAAMjlcoXFryqTy+WIj4/H+fPn0aNHD7zxxhv4/vvvoaHx5LtE3nzzTdjY2CAvLw9jx46tsk1kZCRMTU3Rtm1bDB8+HJcuXcLOnTurjGXChAkwNTVFy5Yt4eXlBTs7Oxw8ePCpcTwPPz8/xMfH4/bt2xg2bBjatWuHiRMnQk9PT0ysu3btil27diE2NhYdOnTAJ598gkWLFtVqxXUdHR0kJSWhVatWGD58OGxtbeHr64uHDx8q3Lf/ovpVZdy4cSgpKcHatWthZmaGEydOoKysDO+88w46duyIGTNmwMDAAGpq1f+T3blzZ6xevRrLli1Dhw4dsH379hrfYlAdExMTJCQk4MKFC/D29n5i4k9ERERE1ZMIz7PCFBERUSV5eXnQ19dHbm5urb+AICKimktLS0NoaChCQ0MbxH3EFfMNCRmEVq2Mn9o+PT0LYWEHatz+RarYdl2/Nw3tmHgZ1OZzD894ExEREREREakQE28iIiIiIiIiFWLiTURERERERKRCTLyJiIiIiF4xpqamCA0NhampaX2HQi8JHhMvN9UtSUxERERERCohlUq5gBYp4DHxcuMZbyIiIiIiIiIVYuJNREREREREpEJMvImIiIiIiIhUiIk3ERERERERkQox8SYiIiIiIiJSIa5qTkREREREr4TMzNxatatp+xepPrZJLz8m3kRERERE9FKTyWTQ0tJCdPSJWvWrbfsXRUtLCzKZrF62TS8niSAIQn0HQUREr4e8vDzo6+sjNzcXenp69R0OERG9RrKyspCfn1/fYdSITCaDsbFxfYdBKlabzz08401ERERERC89Y2NjJrP0yuLiakREREREREQqxMSbiIiIiIiISIWYeBMRERERERGpEBNvIiIiIiIiIhVi4k1ERERERESkQlzVnIiIiIiIXiqv0qPDnhUfOdawMPEmIiIiIqKXRlZWFkJCQlBcXFzfoaiUlpYWwsLCmHw3EEy8iYiIiIjopZGfn4/i4mL06+sGff36T0pzcrOQdHw/HPu6weAFxZObm4Wfju9Hfn4+E+8Ggok3ERERERG9dPT1jdHE2KS+wxAZvGTx0KuFi6sRERERERERqRATbyIiIiIiIiIVYuJNREREREREpEJMvImIiIiIiIhUiIk3ERERERERkQox8SYiIiIioqcqKipCWloaioqK6juUBof7/tXHxJuIiIiIiJ4qIyMDoaGhyMjIqO9QGhzu+1cfE28iIiIiIiIiFWLiTURERERERKRCTLyJ6pFEIsF3332n8u0cO3YMEokEOTk5Kt8WvdrGjx+PoUOH1ncYRERERK8VJt5EKpKZmYnp06fD0tISUqkU5ubmcHd3x9GjR+s8lt69eyMjIwP6+vq17rtnzx6oq6vj9u3bVdZbW1sjKCgIAODk5ASJRAKJRAKpVIoWLVrA3d0de/fuVepX0U4ikUBDQwOtWrVCUFCQyhcNedKXHU/7giI0NBQSiQQDBgxQqluxYgUkEgmcnJyq3XZaWprCvI2MjNCvXz8cP378GWZCRERERK8KJt5EKpCWlgYHBwckJCRgxYoVuHDhAg4dOgRnZ2f4+/vXeTxaWlowMTGBRCKpdd/BgwfD2NgYmzdvVqpLSkpCSkoKfH19xbJJkyYhIyMDqamp2LNnD9q3b49Ro0Zh8uTJSv2jo6ORkZGBGzduYP369di6dSuWLFlS49gqEtm6ZGpqisTERNy6dUuhPCoqCq1atarRGEeOHEFGRgaSkpJgZmYGNzc3/PPPP88cU3Fx8TP3JSIiIiLVY+JNpAJ+fn6QSCT49ddf4eHhAblcDjs7OwQFBeH06dPV9gsODoZcLoeOjg4sLS0xf/58lJSUiPXnz5+Hs7MzZDIZ9PT04ODggLNnzwIAbt68CXd3dxgaGkJXVxd2dnY4cOAAgKrP5J44cQJOTk7Q0dGBoaEhXF1dkZ2drRSTpqYmxowZg5iYGKW6qKgo9OzZE3Z2dmKZjo4OTExM0LJlS/Tq1QvLli3Dpk2bEBkZiSNHjij0NzAwgImJCczNzeHm5oYhQ4bg3LlzNdrH9aVZs2Z45513FL6IOHnyJP777z+8++67NRrD2NgYJiYm6NChA+bNm4e8vDz88ssvYv3FixcxcOBANG7cGM2bN8eYMWPw33//ifVOTk6YNm0aZsyYgSZNmsDV1RUAsHr1anTs2BG6urowNzeHn58fCgoKxH4xMTEwMDDA4cOHYWtri8aNG2PAgAFPXCH1zJkzaNq0KZYtW1bjfUREREREijTqOwCi1829e/dw6NAhLF26FLq6ukr1BgYG1faVyWSIiYmBmZkZLly4gEmTJkEmk2HOnDkAAG9vb9jb22PDhg1QV1dHcnIyNDU1AQD+/v4oLi5GUlISdHV1cenSJTRu3LjK7SQnJ6N///7w8fHBmjVroKGhgcTERJSVlVXZ3tfXF6tXr0ZSUhIcHR0BAAUFBdi9ezfCw8Ofuk/GjRuHmTNnYu/evXBxcamyzdWrV5GQkIDx48c/dbz65uPjgzlz5uCjjz4C8OgLCG9v71qPU1hYiC1btgB4dFUCAOTk5OCtt97CxIkTER4ejsLCQgQHB2PkyJFISEgQ+27evBlTp07FiRMnxDI1NTVERESgTZs2uH79Ovz8/DBnzhysX79ebPPgwQOsXLkSW7duhZqaGt5//33MmjUL27dvV4ovISEBw4cPx/Lly6u8YgF49FzRyrcH5OXl1Xo/EBHRq+XOnTuv9Pgvk5rOtSHtk9cVE2+iFywlJQWCIKBdu3a17vvxxx+Lv1tYWGDWrFmIjY0VE+/09HTMnj1bHNva2lpsn56eDg8PD3Ts2BEAYGlpWe12li9fjm7duikkZJXPWj+uffv26NWrF6KiosTEe9euXRAEAaNGjXrqvNTU1CCXy5GWlqZQ7uXlBXV1dZSWlqKoqAhubm4ICQl56nj1zc3NDVOmTEFSUhIcHBywa9cu/Pzzz4iKiqpR/969e0NNTQ0PHjyAIAhwcHBA//79AQBr166Fvb09Pv30U7F9VFQUzM3NcfXqVcjlcgCP3vvly5crjDtjxgzxdwsLCyxZsgRTpkxReJ9LSkqwceNGtG3bFgAwbdo0LFq0SCnGffv2YezYsfjqq6/w3nvvVTuXsLAwLFy4sEbzJiKi18OXX35Z3yG8NrgvGw4m3kQvmCAIz9x3586diIiIQGpqKgoKClBaWgo9PT2xPigoCBMnTsTWrVvh4uICT09PMYEKCAjA1KlTER8fDxcXF3h4eKBTp05Vbic5ORmenp61is3HxweBgYH44osvIJPJEBUVBU9PT8hkshr1FwRB6X7s8PBwuLi4oKysDCkpKQgKCsKYMWMQGxtb7Th2dna4efOmOCYAhTP7ffv2xcGDB2s1t9rS1NTE+++/j+joaFy/fh1yubzafV2VnTt3ol27drh48SLmzJmDmJgY8cqF8+fPIzExscqrFVJTU8XE28HBQan+yJEjCAsLw5UrV5CXl4fS0lI8fPgQDx48gI6ODoBHtwJUHDPAo3vW7969qzDOL7/8gv3792P37t1PXeE8JCREXFwPeHTG29zcvGY7goiIXkmTJ0+GmZmZysa/c+dOg0lIa7ovG9I+eV0x8SZ6waytrSGRSHDlypVa9Tt16hS8vb2xcOFCuLq6Ql9fH7GxsVi1apXYJjQ0FKNHj0ZcXBwOHjyIBQsWIDY2FsOGDcPEiRPh6uqKuLg4xMfHIywsDKtWrcL06dOVtqWtrV3reY0aNQqBgYHYtWsXHB0dceLECYSFhdWob1lZGa5du4bu3bsrlJuYmMDKygoAYGNjg/z8fHh5eWHJkiVi+eMOHDgg3vd++/ZtODk5ITk5+bnm9ix8fHzQs2dPXLx4ET4+PrXqa25uDmtra1hbW6O0tBTDhg3DxYsXIZVKUVBQAHd39yrvqTY1NRV/f/w2hrS0NLi5uWHq1KlYunQpjIyM8PPPP8PX1xfFxcVi4l2R4FeQSCRKXxa1bdsWxsbGiIqKwrvvvqvUpzKpVAqpVFqr+RMR0avNzMwMFhYW9R3Ga4H7suHg4mpEL5iRkRFcXV2xbt063L9/X6m+ukdVnTx5Eq1bt8ZHH32Ebt26wdraWjyzW5lcLkdgYCDi4+MxfPhwREdHi3Xm5uaYMmUK9u7di5kzZyIyMrLKbXXq1KnWjzWTyWTw9PREVFQUoqOjIZfL0bdv3xr13bx5M7Kzs+Hh4fHEdurq6gAe3ftcndatW8PKygpWVlZo3bo1AIivrays0KJFixrO6PnY2dnBzs4OFy9exOjRo595nBEjRkBDQ0O8HLxr1674888/YWFhoTAvKyurKtcMqPDbb7+hvLwcq1atQq9evSCXy5/5frAmTZogISEBKSkpGDlypMICf0RERERUe0y8iVRg3bp1KCsrQ48ePbBnzx5cu3YNly9fRkREBN54440q+1hbWyM9PR2xsbFITU1FREQE9u3bJ9YXFhZi2rRpOHbsGG7evIkTJ07gzJkzsLW1BfDo/t7Dhw/jxo0bOHfuHBITE8W6x4WEhODMmTPw8/PDH3/8gStXrmDDhg0KK2dXxdfXFydPnsTGjRurPcv74MEDZGZm4tatWzh9+jSCg4MxZcoUTJ06Fc7Ozgptc3JykJmZiTt37uCnn37CokWLIJfLq437Rblx4waSk5MVfip/SXLhwgWFuvPnz1c5TkJCAjIyMp64YN7TSCQSBAQE4LPPPsODBw/g7++Pe/fuwcvLC2fOnEFqaioOHz6MCRMmVLv4HfDoy4eSkhJ88cUXuH79OrZu3YqNGzc+c1zNmjVDQkICrly5Ai8vL5SWlj7zWEREREQNHRNvIhWwtLTEuXPn4OzsjJkzZ6JDhw54++23cfToUWzYsKHKPoMHD0ZgYCCmTZuGLl264OTJk5g/f75Yr66ujqysLIwdOxZyuRwjR47EwIEDxYWtysrK4O/vD1tbWwwYMAByuVxhUa3K5HI54uPjcf78efTo0QNvvPEGvv/+e2hoPPnukzfffBM2NjbIy8vD2LFjq2wTGRkJU1NTtG3bFsOHD8elS5ewc+fOKmOZMGECTE1N0bJlS3h5ecHOzg4HDx58ahzPKygoCPb29go/v//+u1jv6OioUFfV/dTAo8u9nyfprjBu3DiUlJRg7dq1MDMzw4kTJ1BWVoZ33nkHHTt2xIwZM2BgYAA1ter/ye7cuTNWr16NZcuWoUOHDti+fXuNbwWojomJCRISEnDhwgV4e3s/MfEnIiIioupJhOdZCYqIiKiSvLw86OvrIzc3V2FhQCIievWlpaUhNDQUoaGhKr0vuWI7g93GoYmxicq2U1P/ZWXih/2bX2g8FWPWdF/W1b6n2qnN5x6e8SYiIiIiIiJSISbeRERERERERCrExJuIiIiIiIhIhZh4ExERERHRU5mamiI0NBSmpqb1HUqDw33/6lPt0sFERERERPRakEqlXNirnnDfv/p4xpuIiIiIiIhIhZh4ExEREREREakQE28iIiIiIiIiFWLiTURERERERKRCTLyJiIiIiIiIVIirmhMRERER0UsnNzervkMAAOT8/zhyXmA8L8vcqO4w8SYiIiIiopeGTCaDlpYWfjq+v75DUZD0guPR0tKCTCZ7oWPSy0siCIJQ30EQEdHrIS8vD/r6+sjNzYWenl59h0NERK+orKws5Ofn13cYKiWTyWBsbFzfYdBzqM3nHp7xJiIiIiKil4qxsTGTUnqtcHE1IiIiIiIiIhVi4k1ERERERESkQky8iYiIiIiIiFSIiTcRERERERGRCjHxJiIiIiIiIlIhrmpORERERPSSagiP1aotPoaLXkVMvImIiIiIXkJZWVkImTsXxSUl9R3KS0VLUxNhn33G5JteKUy8iYiIiIheQvn5+SguKYFXq+ZoJtWs023ffViMb/6+Cy/zZmjWSKtOt/0kd4tK8E36P8jPz2fiTa8UJt5ERERERC+xZlJNtNRpVD/bbqRVb9smep1wcTUiIiIiIiIiFWLiTURERERERKRCTLyJiIiIiIiIVIiJNxEREREREZEKMfEmIiIiIiIiUiEm3kRERET02ikqKkJaWhqKiorqOxRqoHgMUmVMvImIiIjotZORkYHQ0FBkZGTUdyjUQPEYpMqYeBMRERERERGpEBNvIiIiIiIiIhVi4k1ETyWRSPDdd9+pfDvHjh2DRCJBTk6OyrdFRERERFRXmHgTNXCZmZmYPn06LC0tIZVKYW5uDnd3dxw9erTOY+nduzcyMjKgr69f67579uyBuro6bt++XWW9tbU1goKCAABOTk6QSCSQSCSQSqVo0aIF3N3dsXfvXqV+Fe0kEgk0NDTQqlUrBAUFqXyhlOq+7Bg/fjyGDh1a43FiYmLE+NXV1WFoaIiePXti0aJFyM3NVRq78nwrfgYMGPCcsyEiIiJq2Jh4EzVgaWlpcHBwQEJCAlasWIELFy7g0KFDcHZ2hr+/f53Ho6WlBRMTE0gkklr3HTx4MIyNjbF582aluqSkJKSkpMDX11csmzRpEjIyMpCamoo9e/agffv2GDVqFCZPnqzUPzo6GhkZGbhx4wbWr1+PrVu3YsmSJTWOLS0t7Znm9KLo6ekhIyMDt27dwsmTJzF58mRs2bIFXbp0wZ07dxTaDhgwABkZGQo/33zzTT1FTkRERPR6YOJN1ID5+flBIpHg119/hYeHB+RyOezs7BAUFITTp09X2y84OBhyuRw6OjqwtLTE/PnzUVJSItafP38ezs7OkMlk0NPTg4ODA86ePQsAuHnzJtzd3WFoaAhdXV3Y2dnhwIEDAKq+1PzEiRNwcnKCjo4ODA0N4erqiuzsbKWYNDU1MWbMGMTExCjVRUVFoWfPnrCzsxPLdHR0YGJigpYtW6JXr15YtmwZNm3ahMjISBw5ckShv4GBAUxMTGBubg43NzcMGTIE586dq9E+VqUtW7bA2NhY6ez70KFDMWbMGPG1RCKBiYkJTE1NYWtrC19fX5w8eRIFBQWYM2eOQl+pVAoTExOFH0NDwzqZDxEREdHrSqO+AyCi+nHv3j0cOnQIS5cuha6urlK9gYFBtX1lMhliYmJgZmaGCxcuYNKkSZDJZGIS5+3tDXt7e2zYsAHq6upITk6GpqYmAMDf3x/FxcVISkqCrq4uLl26hMaNG1e5neTkZPTv3x8+Pj5Ys2YNNDQ0kJiYiLKysirb+/r6YvXq1UhKSoKjoyMAoKCgALt370Z4ePhT98m4ceMwc+ZM7N27Fy4uLlW2uXr1KhISEjB+/Pinjqdqnp6eCAgIwA8//ABPT08AwN27dxEXF4f4+Pgn9m3WrBm8vb0RFRWFsrIyqKurP1MMRUVFCol/Xl7eM41DRKQqj1/Z8yp5lWNXtVdh37wKMVLdYeJN1EClpKRAEAS0a9eu1n0//vhj8XcLCwvMmjULsbGxYuKdnp6O2bNni2NbW1uL7dPT0+Hh4YGOHTsCACwtLavdzvLly9GtWzesX79eLKt81vpx7du3R69evRAVFSUm3rt27YIgCBg1atRT56Wmpga5XI60tDSFci8vL6irq6O0tBRFRUVwc3NDSEjIU8d7XhXbrayoqAjvvvsuAEBbWxujR49GdHS0mHhv27YNrVq1gpOT01PHb9euHfLz85GVlYVmzZoBAPbv36/0Rci8efMwb968KscICwvDwoULazs1IqI68+WXX9Z3CKQCfF/pVcPEm6iBEgThmfvu3LkTERERSE1NRUFBAUpLS6GnpyfWBwUFYeLEidi6dStcXFzg6emJtm3bAgACAgIwdepUxMfHw8XFBR4eHujUqVOV20lOThYTypry8fFBYGAgvvjiC8hkMkRFRcHT0xMymaxG/QVBULofOzw8HC4uLigrK0NKSgqCgoIwZswYxMbGVjuOnZ0dbt68KY4JQCGh7du3Lw4ePPjEWCq2W1lwcLDCGf9Jkyahe/fuuH37Nlq0aIGYmBhxkbSazBWAQltnZ2ds2LBBoZ2RkVG1Y4SEhIiL1gGPznibm5s/ddtERHVl8uTJMDMzq+8wnsmdO3eYYFbjVXhf+f5RZUy8iRooa2trSCQSXLlypVb9Tp06BW9vbyxcuBCurq7Q19dHbGwsVq1aJbYJDQ3F6NGjERcXh4MHD2LBggWIjY3FsGHDMHHiRLi6uoqXQ4eFhWHVqlWYPn260ra0tbVrPa9Ro0YhMDAQu3btgqOjI06cOIGwsLAa9S0rK8O1a9fQvXt3hXITExNYWVkBAGxsbJCfnw8vLy8sWbJELH/cgQMHxPveb9++DScnJyQnJ9dqbpW3W0EmkyncA29vb4/OnTtjy5YteOedd/Dnn38iLi6uJtPF5cuXoaenB2NjY7FMV1e32jlVRSqVQiqV1rg9EVFdMzMzg4WFRX2HQS8Y31d61XBxNaIGysjICK6urli3bh3u37+vVF/ds7RPnjyJ1q1b46OPPkK3bt1gbW0tntmtTC6XIzAwEPHx8Rg+fDiio6PFOnNzc0yZMgV79+7FzJkzERkZWeW2OnXqVOvHmslkMnh6eiIqKgrR0dGQy+Xo27dvjfpu3rwZ2dnZ8PDweGK7isu/CwsLq23TunVrWFlZwcrKCq1btwYA8bWVlRVatGhRwxk93cSJExETE4Po6Gi4uLjU6Izz3bt3sWPHDgwdOhRqavxfAREREZEq8Yw3UQO2bt069OnTBz169MCiRYvQqVMnlJaW4scff8SGDRtw+fJlpT7W1tZIT09HbGwsunfvjri4OOzbt0+sLywsxOzZszFixAi0adMGt27dwpkzZ8RkdsaMGRg4cCDkcjmys7ORmJgIW1vbKuMLCQlBx44d4efnhylTpkBLSwuJiYnw9PREkyZNqp2Xr68v+vbti8uXLyM4OLjKNg8ePEBmZiZKS0tx69Yt7Nu3D+Hh4Zg6dSqcnZ0V2ubk5CAzMxPl5eW4du0aFi1aBLlcXm3cdW306NGYNWsWIiMjsWXLFqV6QRCQmZkJQRCQk5ODU6dO4dNPP4W+vj4+++wzhbZFRUXIzMxUKNPQ0Hji/iYiIiKiJ2PiTdSAWVpa4ty5c1i6dClmzpyJjIwMNG3aFA4ODkr3+VYYPHgwAgMDMW3aNHGhr/nz5yM0NBTAo7PBWVlZGDt2LP755x80adIEw4cPFxfgKisrg7+/P27dugU9PT0MGDCg2hXH5XI54uPjMW/ePPTo0QPa2tro2bMnvLy8njivN998EzY2NkhJScHYsWOrbBMZGYnIyEhoaWnB2NgYDg4O2LlzJ4YNG6bUdsKECQD+77Fcjo6O+PTTT6Gh8XL8E6qvrw8PDw/ExcVh6NChSvV5eXkwNTWFRCKBnp4ebGxsMG7cOHz44YcK9+YDwKFDh2BqaqpQZmNjU+tbEoiIiIjo/0iE51lhiYiIXgr9+/eHnZ0dIiIi6jWOvLw86OvrIzc3VympJyKqS2lpaQgNDUVoaOgrey9wxRw+tG6JljqN6nTbtx48xJprt+pl209SEder8L6+DscgPVltPve8HKdriIjomWRnZ+PYsWM4duyYwmPXiIiIiOjlwcSbiOgVZm9vj+zsbCxbtgw2Njb1HQ4RERERVYGJNxHRKywtLa2+QyAiIiKip+AzZIiIiIjotWNqaorQ0FClBSOJ6gqPQaqMZ7yJiIiI6LUjlUq5oBXVKx6DVBnPeBMRERERERGpEBNvIiIiIiIiIhVi4k1ERERERESkQky8iYiIiIiIiFSIiTcRERERERGRCnFVcyIiIiKil9jdopK63+bDYoX/vizqY18QvQhMvImIiIiIXkIymQxampr4Jv2feovhm7/v1tu2q6OlqQmZTFbfYRDVikQQBKG+gyAiotdDXl4e9PX1kZubCz09vfoOh4jolZeVlYX8/Pz6DuOlIpPJYGxsXN9hENXqcw/PeBMRERERvaSMjY2ZZBK9Bri4GhEREREREZEKMfEmIiIiIiIiUiEm3kREREREREQqxMSbiIiIiIiISIWYeBMRERERERGpEFc1JyIiIiIiek6qfvQbH6P2amPiTURERERE9ByysrIwd24ISkqKVbYNTU0tfPZZGJPvVxQTbyIiIiIioueQn5+PkpJitHBwgVRmWON+RfnZuP3bkaf2q2iXn5/PxPsVxcSbiIiIiIjoBZDKDKFt0LTO+tGrg4urEREREREREakQE28iIiIiIiIiFWLiTURERERERKRCTLyJiIiIiIiIVIiJNxEREREREZEKMfEmIiIiIqLXWlFREdLS0lBUVFTfoahMQ5jjq4yJNxERERERvdYyMjIQGhqKjIyM+g5FZRrCHF9lTLyJiIiIiIiIVIiJNxEREREREZEKMfGmZyaRSPDdd9+pfDvHjh2DRCJBTk6OyrdFr5eYmBgYGBjUdxivjLS0NEgkEiQnJ9d3KERERETPpLy8HJcvX8bp06dx+fJllJeX13dIAJh4UzUyMzMxffp0WFpaQiqVwtzcHO7u7jh69Gidx9K7d29kZGRAX1+/1n337NkDdXV13L59u8p6a2trBAUFAQCcnJwgkUggkUgglUrRokULuLu7Y+/evUr9KtpJJBJoaGigVatWCAoKUvliFtV92TF+/HgMHTq0xuPExMSI8aurq8PQ0BA9e/bEokWLkJubqzR25flW/AwYMOA5Z/NkeXl5+Oijj9CuXTs0atQIJiYmcHFxwd69eyEIgkq3rSqV96WmpibatGmDOXPm4OHDh/UdGhEREdEr7+zZs5gzZw6WLVuGjRs3YtmyZZgzZw7Onj1b36Ex8SZlaWlpcHBwQEJCAlasWIELFy7g0KFDcHZ2hr+/f53Ho6WlBRMTE0gkklr3HTx4MIyNjbF582aluqSkJKSkpMDX11csmzRpEjIyMpCamoo9e/agffv2GDVqFCZPnqzUPzo6GhkZGbhx4wbWr1+PrVu3YsmSJTWOreLsYn3R09NDRkYGbt26hZMnT2Ly5MnYsmULunTpgjt37ii0HTBgADIyMhR+vvnmmxpva/z48QgNDa1x+5ycHPTu3RtbtmxBSEgIzp07h6SkJLz33nuYM2eO0pcDda2kpOSZ+1bsy+vXryM8PBybNm3CggULnnk8QRBQWlr6zP2JiIiIXgdnz57FunXr0LJlS3z88cfYuHEjPv74Y7Rs2RLr1q2r9+SbiTcp8fPzg0Qiwa+//goPDw/I5XLY2dkhKCgIp0+frrZfcHAw5HI5dHR0YGlpifnz5yskKOfPn4ezszNkMhn09PTg4OAg/gHcvHkT7u7uMDQ0hK6uLuzs7HDgwAEAVV9qfuLECTg5OUFHRweGhoZwdXVFdna2UkyampoYM2YMYmJilOqioqLQs2dP2NnZiWU6OjowMTFBy5Yt0atXLyxbtgybNm1CZGQkjhw5otDfwMAAJiYmMDc3h5ubG4YMGYJz587VaB+r0pYtW2BsbKx09n3o0KEYM2aM+FoikcDExASmpqawtbWFr68vTp48iYKCAsyZM0ehr1QqhYmJicKPoaGhyuYwb948pKWl4ZdffsG4cePQvn17yOVyTJo0CcnJyWjcuDEAIDs7G2PHjoWhoSF0dHQwcOBAXLt2TWm8w4cPw9bWFo0bNxYT38q++uor2NraolGjRmjXrh3Wr18v1lV8QbJz507069cPjRo1wvbt25/arzoV+9Lc3BxDhw6Fi4sLfvzxR7G+vLwcYWFhaNOmDbS1tdG5c2fs3r1brK/4ezh48CAcHBwglUrx888/IzU1FUOGDEHz5s3RuHFjdO/eXemYtbCwwKeffgofHx/IZDK0atUKX375ZbWxlpWVwcfHB+3atUN6evpT50ZERERUH8rLyxEbG4vOnTsjICAAVlZWaNSoEaysrBAQEIDOnTtj586d9XrZuUa9bZleSvfu3cOhQ4ewdOlS6OrqKtU/6X5ZmUyGmJgYmJmZ4cKFC5g0aRJkMpmYxHl7e8Pe3h4bNmyAuro6kpOToampCQDw9/dHcXExkpKSoKuri0uXLonJ1eOSk5PRv39/+Pj4YM2aNdDQ0EBiYiLKysqqbO/r64vVq1cjKSkJjo6OAICCggLs3r0b4eHhT90n48aNw8yZM7F37164uLhU2ebq1atISEjA+PHjnzqeqnl6eiIgIAA//PADPD09AQB3795FXFwc4uPjn9i3WbNm8Pb2RlRUFMrKyqCurl4XISuo+IfT29sbZmZmSvWVj4vx48fj2rVr+OGHH6Cnp4fg4GAMGjQIly5dEo+tBw8eYOXKldi6dSvU1NTw/vvvY9asWWLyvH37dnzyySdYu3Yt7O3t8fvvv2PSpEnQ1dXFuHHjxG3NnTsXq1atgr29vZh816Tfk1y8eBEnT55E69atxbKwsDBs27YNGzduhLW1NZKSkvD++++jadOm6Nevn0I8K1euhKWlJQwNDfH3339j0KBBWLp0KaRSKbZs2QJ3d3f89ddfaNWqldhv1apVWLx4MebNm4fdu3dj6tSp6NevH2xsbBRiKyoqgpeXF9LS0nD8+HE0bdq0yjkUFRUpfMmTl5dXo7kTERHVh8ev6nvZx63NduoqhpfRX3/9hf/++w9TpkyBmpriuWU1NTW4ublhyZIl+Ouvv2Bra1svMTLxJgUpKSkQBAHt2rWrdd+PP/5Y/N3CwgKzZs1CbGysmHinp6dj9uzZ4tjW1tZi+/T0dHh4eKBjx44AAEtLy2q3s3z5cnTr1k3h7GLls9aPa9++PXr16oWoqCgx8d61axcEQcCoUaOeOi81NTXI5XKkpaUplHt5eUFdXR2lpaUoKiqCm5sbQkJCnjre86rYbmVFRUV49913AQDa2toYPXo0oqOjxcR727ZtaNWqFZycnJ46frt27ZCfn4+srCw0a9YMALB//36lL0LmzZuHefPmvYAZKfrvv/+QnZ391GOwIuE+ceIEevfuDeBREm1ubo7vvvtOnHtJSQk2btyItm3bAgCmTZuGRYsWieMsWLAAq1atwvDhwwEAbdq0waVLl7Bp0yaFBHrGjBlim9r0e1zFvqw4btTU1LB27VoAj97HTz/9FEeOHMEbb7wB4NHfws8//4xNmzYpJN6LFi3C22+/Lb42MjJC586dxdeLFy/Gvn378MMPP2DatGli+aBBg+Dn5wfg0VUq4eHhSExMVEi8CwoK8O6776KoqAiJiYlPXF8hLCwMCxcurLaeiIjoZfKkK71eBa96/KpScRtiy5Ytq6xv0aKFQrv6wMSbFDzPolU7d+5EREQEUlNTUVBQgNLSUujp6Yn1QUFBmDhxIrZu3QoXFxd4enqKyVBAQACmTp2K+Ph4uLi4wMPDA506dapyO8nJyWJSVVM+Pj4IDAzEF198AZlMhqioKHh6ekImk9WovyAISvdjh4eHw8XFBWVlZUhJSUFQUBDGjBmD2NjYasexs7PDzZs3xTEBxTO4ffv2xcGDB58YS8V2KwsODlY44z9p0iR0794dt2/fRosWLRATEyMu7FWTuQJQaOvs7IwNGzYotDMyMqp2jO3bt+ODDz4QXxcVFUEikWDlypVi2cGDB9G3b99qt/80ly9fhoaGBnr27CmWGRsbw8bGBpcvXxbLdHR0xOMMAExNTXH37l0AwP3795GamgpfX19MmjRJbFNaWqqUbHbr1k38vTb9HlexL+/fv4/w8HBoaGjAw8MDwKMvvh48eKCQUANAcXEx7O3tq40HeJQsh4aGIi4uDhkZGSgtLUVhYaHSJeKV/64qbjeo2B8VvLy80LJlSyQkJEBbW/uJ8wkJCREXKAQenfE2Nzd/Yh8iIqL6Mnny5CqvqHted+7cqZOk+Enx11UML6OKz1+3bt2ClZWVUn3FQsvPsljzi8LEmxRYW1tDIpHgypUrtep36tQpeHt7Y+HChXB1dYW+vj5iY2OxatUqsU1oaChGjx6NuLg4HDx4EAsWLEBsbCyGDRuGiRMnwtXVVbwcOiwsDKtWrcL06dOVtvW0RKAqo0aNQmBgIHbt2gVHR0ecOHECYWFhNepbVlaGa9euoXv37grlJiYm4h+2jY0N8vPz4eXlhSVLllT5Bw8ABw4cEO97v337NpycnBQe3VSTuVXebgWZTKZwD7y9vT06d+6MLVu24J133sGff/6JuLi4mkwXly9fhp6eHoyNjcUyXV3daudUlcGDByskxMHBwWjRogUCAgLEsopvHh/XtGlTGBgY1PoYrE7FJecVJBKJmNwXFBQAACIjIxXiBaB0VUHlWy9q0+9xlfdlVFQUOnfujK+//hq+vr7iuHFxcUr7RyqVVhsPAMyaNQs//vgjVq5cCSsrK2hra2PEiBEoLi5WaFfV/nj8fqdBgwZh27ZtOHXqFN56660nzkcqlSrFRkRE9LIyMzODhYVFfYfxzF71+FXFxsYGTZo0wf79+xEQEKBwuXl5eTn279+Ppk2bKt1aV5eYeJMCIyMjuLq6Yt26dQgICFD6cJ+Tk1Plfd4V96l+9NFHYlnFmd3K5HI55HI5AgMD4eXlhejoaAwbNgwAYG5ujilTpmDKlCkICQlBZGRklYl3p06dcPTo0Vpd3iqTyeDp6YmoqCikpqZCLpdXeba1Kps3b0Z2drZ4VrI6FQlXYWFhtW0q38urofHoz682CW1tTJw4EZ9//jlu374NFxeXGp2FvHv3Lnbs2IGhQ4cq3R9TGzKZTOFqAplMBiMjoxrNVU1NDaNGjcLWrVuxYMECpW91CwoK0KhRI9ja2qK0tBS//PKLeKl5VlYW/vrrL7Rv375GcTZv3hxmZma4fv06vL29azy/Z+33ODU1NcybNw9BQUEYPXo02rdvD6lUivT0dIXLymvixIkTGD9+vPj3VFBQoHR7RE1NnToVHTp0wODBgxEXF1frWIiIiIjqUsXnx3Xr1iEiIgJubm5o0aIFbt++jf379+P8+fPw9/d/rs+3z4uJNylZt24d+vTpgx49emDRokXo1KkTSktL8eOPP2LDhg0Kl/FWsLa2Rnp6OmJjY9G9e3fExcVh3759Yn1hYSFmz56NESNGoE2bNrh16xbOnDkjJrMzZszAwIEDIZfLkZ2djcTExGoXPggJCUHHjh3h5+eHKVOmQEtLC4mJifD09ESTJk2qnZevry/69u2Ly5cvIzg4uMo2Dx48QGZmJkpLS3Hr1i3s27cP4eHhmDp1KpydnRXa5uTkIDMzE+Xl5bh27RoWLVoEuVxebws2PG706NGYNWsWIiMjsWXLFqV6QRCQmZkJQRCQk5ODU6dO4dNPP4W+vj4+++wzhbZFRUXIzMxUKNPQ0Hji/n4eS5cuxbFjx9CzZ08sXboU3bp1g6amJo4fP46wsDCcOXMG1tbWGDJkCCZNmoRNmzZBJpNh7ty5aNGiBYYMGVLjbS1cuBABAQHQ19fHgAEDUFRUhLNnzyI7O1vhEuoX1e9xnp6emD17NtatW4dZs2Zh1qxZCAwMRHl5Od58803k5ubixIkT0NPTe+K949bW1ti7dy/c3d0hkUgwf/7851q5c/r06SgrK4ObmxsOHjyIN99885nHIiIiIlK1bt26wd/fH7GxsQqP+G3atCn8/f2VbtOra0y8SYmlpSXOnTuHpUuXYubMmcjIyEDTpk3h4OCgdJ9vhcGDByMwMBDTpk0TF/qaP3+++OxmdXV1ZGVlYezYsfjnn3/QpEkTDB8+XDxrXVZWBn9/f9y6dQt6enoYMGBAtSuOy+VyxMfHY968eejRowe0tbXRs2dPeHl5PXFeb775JmxsbJCSkoKxY8dW2SYyMhKRkZHQ0tKCsbExHBwcsHPnTvEsYmUTJkwA8H/3yTo6OuLTTz8Vz2TXN319fXh4eCAuLg5Dhw5Vqs/Ly4OpqSkkEgn09PRgY2ODcePG4cMPP1S4Nx8ADh06BFNTU4UyGxubF3Y5+OOMjIxw+vRpfPbZZ1iyZAlu3rwJQ0NDdOzYEStWrBDvz4mOjsaHH34INzc3FBcXw9HREQcOHFC6nPpJJk6cCB0dHaxYsQKzZ8+Grq4uOnbsiBkzZqik3+M0NDQwbdo0LF++HFOnTsXixYvRtGlThIWF4fr16zAwMEDXrl2fupDd6tWr4ePjg969e6NJkyYIDg5+7hXGZ8yYgfLycgwaNAiHDh0SrywgIiIiehl169YNXbt2xV9//YXc3Fzo6+vDxsamXs90V5AIz7OaFhG91Pr37w87OztERETUdyjUQOTl5UFfXx+5ublKX+AQERHVl7S0NISGhiI0NFQl90hXjG/p5Altg6ofwVmVwpx/cf3Yt0/tV9HuSfGreo6krDafe16OU3NE9EJlZ2fj2LFjOHbsmMJj14iIiIiIqO4x8SZ6Ddnb2yM7OxvLli2r19UbiYiIiIiIiTfRa+lZV7MmIiIiIqIXr/7vMiciIiIiIlIhU1NThIaGKi0W+zppCHN8lfGMNxERERERvdakUulrv+BYQ5jjq4xnvImIiIiIiIhUiIk3ERERERERkQox8SYiIiIiIiJSISbeRERERERERCrExJuIiIiIiIhIhbiqORERERER0QtQlJ/9TO2f1q+249LLh4k3ERERERHRc5DJZNDU1MLt3448U/+a9NPU1IJMJnum8an+SQRBEOo7CCIiej3k5eVBX18fubm50NPTq+9wiIiI6kxWVhby8/NVNr5MJoOxsbHKxqfaq83nHp7xJiIiIiIiek7GxsZMjKlaXFyNiIiIiIiISIWYeBMRERERERGpEBNvIiIiIiIiIhVi4k1ERERERESkQky8iYiIiIiIiFSIq5oTEREREVG9UfVjuF42fCxYw8TEm4iIiIiI6kVWVhZC5s5FcUlJfYdSZ7Q0NRH22WdMvhsYJt5ERERERFQv8vPzUVxSAq8eemgmU39h497NK8U3Z/Lh1V2GZnovT8pzN78M3/yah/z8fCbeDczLcxQSEREREVGD1EymjpaGmi9+XD0NlYxLVFtcXI2IiIiIiIhIhZh4ExEREREREakQE28iIiIiIiIiFWLiTURERERERKRCTLyJiIiIiIiIVIiJNxERERERvRBFRUVIS0tDUVFRfYfS4HDfv9yYeBMRERER0QuRkZGB0NBQZGRk1HcoDQ73/cuNiTcRERERERGRCjHxJiIiIiIiIlIhJt700pFIJPjuu+9Uvp1jx45BIpEgJydH5dsiehWkpaVBIpEgOTm5vkMhIiIieq0w8aY6lZmZienTp8PS0hJSqRTm5uZwd3fH0aNH6zyW3r17IyMjA/r6+rXuu2fPHqirq+P27dtV1ltbWyMoKAgA4OTkBIlEAolEAqlUihYtWsDd3R179+5V6lfRTiKRQENDA61atUJQUJDKF8mo7suO8ePHY+jQoTUeJyYmRoxfXV0dhoaG6NmzJxYtWoTc3FylsSvPt+JnwIABzzmb6oWGhqJLly7V1js5OWHGjBnV1lfEePr0aYXyoqIiGBsbQyKR4NixY9X2rzxnTU1NtGnTBnPmzMHDhw9rORMiIiIiepUw8aY6k5aWBgcHByQkJGDFihW4cOECDh06BGdnZ/j7+9d5PFpaWjAxMYFEIql138GDB8PY2BibN29WqktKSkJKSgp8fX3FskmTJiEjIwOpqanYs2cP2rdvj1GjRmHy5MlK/aOjo5GRkYEbN25g/fr12Lp1K5YsWVLj2CrOWtYXPT09ZGRk4NatWzh58iQmT56MLVu2oEuXLrhz545C2wEDBiAjI0Ph55tvvqnxtsaPH4/Q0NAXPIMnMzc3R3R0tELZvn370Lhx4xr1r5jz9evXER4ejk2bNmHBggXPHI8gCCgtLX3m/kRERESkeky8qc74+flBIpHg119/hYeHB+RyOezs7BAUFKR0BrGy4OBgyOVy6OjowNLSEvPnz0dJSYlYf/78eTg7O0Mmk0FPTw8ODg44e/YsAODmzZtwd3eHoaEhdHV1YWdnhwMHDgCo+lLzEydOwMnJCTo6OjA0NISrqyuys7OVYtLU1MSYMWMQExOjVBcVFYWePXvCzs5OLNPR0YGJiQlatmyJXr16YdmyZdi0aRMiIyNx5MgRhf4GBgYwMTGBubk53NzcMGTIEJw7d65G+1iVtmzZAmNjY6Wz70OHDsWYMWPE1xKJBCYmJjA1NYWtrS18fX1x8uRJFBQUYM6cOQp9pVIpTExMFH4MDQ3rZD7Paty4cYiNjUVhYaFYFhUVhXHjxtWof8Wczc3NMXToULi4uODHH38U68vLyxEWFoY2bdpAW1sbnTt3xu7du8X6iuP24MGDcHBwgFQqxc8//4zU1FQMGTIEzZs3R+PGjdG9e3elY8vCwgKffvopfHx8IJPJ0KpVK3z55ZfVxlpWVgYfHx+0a9cO6enpNd1FRERERPQYjfoOgBqGe/fu4dChQ1i6dCl0dXWV6g0MDKrtK5PJEBMTAzMzM1y4cAGTJk2CTCYTkzhvb2/Y29tjw4YNUFdXR3JyMjQ1NQEA/v7+KC4uRlJSEnR1dXHp0qVqz0wmJyejf//+8PHxwZo1a6ChoYHExESUlZVV2d7X1xerV69GUlISHB0dAQAFBQXYvXs3wsPDn7pPxo0bh5kzZ2Lv3r1wcXGpss3Vq1eRkJCA8ePHP3U8VfP09ERAQAB++OEHeHp6AgDu3r2LuLg4xMfHP7Fvs2bN4O3tjaioKJSVlUFdXb0uQlYJBwcHWFhYYM+ePXj//feRnp6OpKQkrFu3DosXL67VWBcvXsTJkyfRunVrsSwsLAzbtm3Dxo0bYW1tjaSkJLz//vto2rQp+vXrJ7abO3cuVq5cCUtLSxgaGuLvv//GoEGDsHTpUkilUmzZsgXu7u7466+/0KpVK7HfqlWrsHjxYsybNw+7d+/G1KlT0a9fP9jY2CjEVlRUBC8vL6SlpeH48eNo2rRplXMoKipS+DImLy+vVvuAiIheT49f5fa87V43qph3Q92Xrwom3lQnUlJSIAgC2rVrV+u+H3/8sfi7hYUFZs2ahdjYWDHxTk9Px+zZs8Wxra2txfbp6enw8PBAx44dAQCWlpbVbmf58uXo1q0b1q9fL5ZVPmv9uPbt26NXr16IiooSE+9du3ZBEASMGjXqqfNSU1ODXC5HWlqaQrmXlxfU1dVRWlqKoqIiuLm5ISQk5KnjPa+K7VZWVFSEd999FwCgra2N0aNHIzo6Wky8t23bhlatWsHJyemp47dr1w75+fnIyspCs2bNAAD79+9X+iJk3rx5mDdv3guYker4+PggKioK77//PmJiYjBo0KBqE9PHVcy54v1VU1PD2rVrATza359++imOHDmCN954A8CjY/bnn3/Gpk2bFBLvRYsW4e233xZfGxkZoXPnzuLrxYsXY9++ffjhhx8wbdo0sXzQoEHw8/MD8OhqkvDwcCQmJiok3gUFBXj33XdRVFSExMTEJ66DEBYWhoULF9Zo7kRE1HA86Yoq4v5piJh4U50QBOGZ++7cuRMRERFITU1FQUEBSktLoaenJ9YHBQVh4sSJ2Lp1K1xcXODp6Ym2bdsCAAICAjB16lTEx8fDxcUFHh4e6NSpU5XbSU5OFhPKmvLx8UFgYCC++OILyGQyREVFwdPTEzKZrEb9BUFQuh87PDwcLi4uKCsrQ0pKCoKCgjBmzBjExsZWO46dnR1u3rwpjglAIaHt27cvDh48+MRYKrZbWXBwsMIZ/0mTJqF79+64ffs2WrRogZiYGHHBsJrMFYBCW2dnZ2zYsEGhnZGRUbVjbN++HR988IH4uqioCBKJBCtXrhTLDh48iL59+z41nufx/vvvY+7cubh+/TpiYmIQERFR474Vc75//z7Cw8OhoaEBDw8PAI++oHrw4IFCQg0AxcXFsLe3Vyjr1q2bwuuCggKEhoYiLi4OGRkZKC0tRWFhodIl4pWP/4rbAu7evavQxsvLCy1btkRCQgK0tbWfOJ+QkBBxIUHg0Rlvc3Pzp+wFIiJ63U2ePBlmZmZPbXfnzp0GmYTWdP/URkPdl68KJt5UJ6ytrSGRSHDlypVa9Tt16hS8vb2xcOFCuLq6Ql9fH7GxsVi1apXYJjQ0FKNHj0ZcXBwOHjyIBQsWIDY2FsOGDcPEiRPh6uoqXg4dFhaGVatWYfr06UrbelqCUZVRo0YhMDAQu3btgqOjI06cOIGwsLAa9S0rK8O1a9fQvXt3hXITExNYWVkBAGxsbJCfnw8vLy8sWbJELH/cgQMHxPveb9++DScnJ4VHQtVkbpW3W0EmkyncA29vb4/OnTtjy5YteOedd/Dnn38iLi6uJtPF5cuXoaenB2NjY7FMV1e32jlVZfDgwejZs6f4Ojg4GC1atEBAQIBY1qJFixqP96yMjY3h5uYGX19fPHz4EAMHDkR+fn6N+laec1RUFDp37oyvv/4avr6+KCgoAADExcUpzUMqlSqNU9msWbPw448/YuXKlbCysoK2tjZGjBiB4uJihXYVt2FUkEgkKC8vVygbNGgQtm3bhlOnTuGtt9564nykUqlSbERERGZmZrCwsKjvMF5a3D8NDxNvqhNGRkZwdXXFunXrEBAQoJQ05OTkVHmfd8X9rx999JFYVnFmtzK5XA65XI7AwEB4eXkhOjoaw4YNA/BoFeopU6ZgypQpCAkJQWRkZJWJd6dOnXD06NFaXTYrk8ng6emJqKgopKamQi6X1/hs6+bNm5GdnS2e7axOxeXflRfzelzle4Q1NB79Wdcmoa2NiRMn4vPPP8ft27fh4uJSo7Obd+/exY4dOzB06FCoqT37mo4ymUzhagKZTAYjIyOVzfVJfHx8MGjQIAQHBz/zPetqamqYN28egoKCMHr0aLRv3x5SqRTp6ekKl5XXxIkTJzB+/HjxuC8oKFC6jaGmpk6dig4dOmDw4MGIi4urdSxEREREpIiJN9WZdevWoU+fPujRowcWLVqETp06obS0FD/++CM2bNiAy5cvK/WxtrZGeno6YmNj0b17d8TFxWHfvn1ifWFhIWbPno0RI0agTZs2uHXrFs6cOSMmszNmzMDAgQMhl8uRnZ2NxMRE2NraVhlfSEgIOnbsCD8/P0yZMgVaWlpITEyEp6cnmjRpUu28fH190bdvX1y+fBnBwcFVtnnw4AEyMzNRWlqKW7duYd++fQgPD8fUqVPh7Oys0DYnJweZmZkoLy/HtWvXsGjRIsjl8mrjrmujR4/GrFmzEBkZiS1btijVC4KAzMxMCIKAnJwcnDp1Cp9++in09fXx2WefKbQtKipCZmamQpmGhsYT9/fzKiwsVLgaAHiUwFfcnvDvv/8q1ZuamqJ58+YKZQMGDMC///6rcNvDs/D09MTs2bOxbt06zJo1C7NmzUJgYCDKy8vx5ptvIjc3FydOnICent4TV063trbG3r174e7uDolEgvnz5yudya6N6dOno6ysDG5ubjh48CDefPPNZx6LiIiIqKFj4k11xtLSEufOncPSpUsxc+ZMZGRkoGnTpnBwcFC6z7fC4MGDERgYiGnTpokLfc2fP198drO6ujqysrIwduxY/PPPP2jSpAmGDx8unrUuKyuDv78/bt26BT09PQwYMKDaFcflcjni4+Mxb9489OjRA9ra2ujZsye8vLyeOK8333wTNjY2SElJwdixY6tsExkZicjISGhpacHY2BgODg7YuXOneHaysgkTJgD4v/tvHR0d8emnn4pnsuubvr4+PDw8EBcXh6FDhyrV5+XlwdTUFBKJBHp6erCxscG4cePw4YcfKiWphw4dgqmpqUKZjY1NrW9JqI2rV68q3S/dv39/8dFbO3bswI4dOxTqFy9erLDIH/Do/XkRXxBoaGhg2rRpWL58OaZOnYrFixejadOmCAsLw/Xr12FgYICuXbs+dcG51atXw8fHB71790aTJk0QHBz83CuMz5gxA+Xl5Rg0aBAOHTqE3r17P9d4RERERA2VRHieVa+IqEHq378/7OzsarWoGDUMeXl50NfXR25u7nNfDUBERK+etLQ0hIaGIjQ0tEb3MFe0/7C/IVoaaj61fU3dyi7BmqPZL3zc51URV033T23Udt/T86vN556X4xQaEb0SsrOzcezYMRw7dkzhsWtERERERFQ9Jt5EVGP29vbIzs7GsmXLFJ77TERERERE1WPiTUQ19qyrZBMRERERNWTP/lwfIiIiIiKiSkxNTREaGqq0eCqpHvf9y41nvImIiIiI6IWQSqVc2KuecN+/3HjGm4iIiIiIiEiFmHgTERERERERqRATbyIiIiIiIiIVYuJNREREREREpEJMvImIiIiIiIhUiKuaExERERFRvbqbX/Zix8srVfjvy+JFz5NeHUy8iYiIiIioXshkMmhpauKbX/NUMv43Z/JVMu7z0NLUhEwmq+8wqI5JBEEQ6jsIIiJ6PeTl5UFfXx+5ubnQ09Or73CIiOgVkJWVhfz8ly9BVhWZTAZjY+P6DoNegNp87uEZbyIiIiIiqjfGxsZMROm1x8XViIiIiIiIiFSIiTcRERERERGRCjHxJiIiIiIiIlIhJt5EREREREREKsTEm4iIiIiIiEiFuKo5ERERERHVyuv8CDA+7otUgYk3ERERERHVWFZWFuaGzEVJcUl9h6ISmlqa+CzsMybf9EIx8SYiIiIiohrLz89HSXEJHAYYQGb09HQi/14pfjuUU+P29aki1vz8fCbe9EK93Ec+ERERERG9lGRGGjBopqmy9kSvEy6uRkRERERERKRCTLyJiIiIiIiIVIiJNxEREREREZEKMfEmIiIiIiIiUiEm3kREREREREQqxMSbiIiIiKgBKyoqQlpaGoqKiuo7FHpJ8Jh48Zh4ExERERE1YBkZGQgNDUVGRkZ9h0IvCR4TLx4TbyIiIiIiIiIVYuJNREREREREpEKvfeItkUjw3XffqXw7x44dg0QiQU5Ojsq3RYrqat+/au+xKo59JycnzJgx44WOWZW6+rt9WTg6OmLHjh31HYaCQ4cOoUuXLigvL6/vUIiIiIheea904p2ZmYnp06fD0tISUqkU5ubmcHd3x9GjR+s8lt69eyMjIwP6+vq17rtnzx6oq6vj9u3bVdZbW1sjKCgIwKPERyKRQCKRQCqVokWLFnB3d8fevXuV+lW0k0gk0NDQQKtWrRAUFKTyRRIqtnn69GmF8qKiIhgbG0MikeDYsWMvbHvPs+9fNAsLC0gkEsTGxirV2dnZQSKRICYmpk5iycjIwMCBA+tkWxViYmLE919NTQ2mpqZ47733kJ6eXqdxvEp++OEH/PPPPxg1apRYVtvj6FnbSyQS6OjooGPHjvjqq68U+g0YMACamprYvn3780+SiIiIqIF7ZRPvtLQ0ODg4ICEhAStWrMCFCxdw6NAhODs7w9/fv87j0dLSgomJCSQSSa37Dh48GMbGxti8ebNSXVJSElJSUuDr6yuWTZo0CRkZGUhNTcWePXvQvn17jBo1CpMnT1bqHx0djYyMDNy4cQPr16/H1q1bsWTJkhrHlpaW9kxzMjc3R3R0tELZvn370Lhx41qP9TTPs+9Voaq5nz59GpmZmdDV1X2uscvKymp8BtLExARSqfS5tvcs9PT0kJGRgdu3b2PPnj3466+/4OnpWedxvCoiIiIwYcIEqKkp/nNc2+Ootu0XLVqEjIwMXLx4Ee+//z4mTZqEgwcPKrQZP348IiIinnVqRERERPT/vbKJt5+fHyQSCX799Vd4eHhALpfDzs4OQUFBSmdaKwsODoZcLoeOjg4sLS0xf/58lJSUiPXnz5+Hs7MzZDIZ9PT04ODggLNnzwIAbt68CXd3dxgaGkJXVxd2dnY4cOAAgKovQz5x4gScnJygo6MDQ0NDuLq6Ijs7WykmTU1NjBkzpsozoVFRUejZsyfs7OzEMh0dHZiYmKBly5bo1asXli1bhk2bNiEyMhJHjhxR6G9gYAATExOYm5vDzc0NQ4YMwblz52q0j5/HuHHjEBsbi8LCQoW5jBs3Tqntk94TQRDg4uICV1dXCIIAALh37x5atmyJTz75BIDyvo+JiYGBgQH2798PGxsb6OjoYMSIEXjw4AE2b94MCwsLGBoaIiAgAGVlZWIcW7duRbdu3SCTyWBiYoLRo0fj7t27tZ67t7c3fvrpJ/z9998Kc/f29oaGhoZC29WrV6Njx47Q1dWFubk5/Pz8UFBQINZXzOWHH35A+/btIZVKkZ6ejoyMDLz77rvQ1tZGmzZtsGPHDlhYWODzzz8X+1a+XLviC5S9e/fC2dkZOjo66Ny5M06dOiW2z8rKgpeXF1q0aCGeBf3mm29qPX+JRAITExOYmpqid+/e8PX1xa+//oq8vDyxzffff4+uXbuiUaNGsLS0xMKFC1FaWlrtmH///TdGjhwJAwMDGBkZYciQIUhLSwMAxMfHo1GjRkq3AHz44Yd46623ajw3JycnBAQEYM6cOTAyMoKJiQlCQ0MV2uTk5OCDDz5A8+bN0ahRI3To0AH79+8X63/++Wf07dsX2traMDc3R0BAAO7fv1/tvP79918kJCTA3d1dqa42x9GztK84zi0tLREcHAwjIyP8+OOPCm3c3d1x9uxZpKamVjsHIiIiIno65U9jr4B79+7h0KFDWLp0aZVncgwMDKrtK5PJEBMTAzMzM1y4cAGTJk2CTCbDnDlzADz68Gpvb48NGzZAXV0dycnJ0NTUBAD4+/ujuLgYSUlJ0NXVxaVLl6o9g5ucnIz+/fvDx8cHa9asgYaGBhITExUSvcp8fX2xevVqJCUlwdHREQBQUFCA3bt3Izw8/Kn7ZNy4cZg5cyb27t0LFxeXKttcvXoVCQkJGD9+/FPHe14ODg6wsLDAnj178P777yM9PR1JSUlYt24dFi9erND2Se+JRCLB5s2b0bFjR0RERODDDz/ElClT0KJFCzHxrsqDBw8QERGB2NhY5OfnY/jw4Rg2bBgMDAxw4MABXL9+HR4eHujTpw/ee+89AEBJSQkWL14MGxsb3L17F0FBQRg/frz45UpNNW/eHK6urti8eTM+/vhjPHjwADt37sRPP/2ELVu2KLRVU1NDREQE2rRpg+vXr8PPzw9z5szB+vXrFeaybNkyfPXVVzA2NkazZs0wZMgQ/Pfffzh27Bg0NTURFBRUoy8JPvroI6xcuRLW1tb46KOP4OXlhZSUFGhoaODhw4dwcHBAcHAw9PT0EBcXhzFjxqBt27bo0aNHrfZBhbt372Lfvn1QV1eHuro6AOD48eMYO3YsIiIi0LdvX6SmpopXayxYsEBpjJKSEri6uuKNN97A8ePHoaGhgSVLlmDAgAH4448/0L9/fxgYGGDPnj3ilSFlZWXYuXMnli5dCgA1ntvmzZsRFBSEX375BadOncL48ePRp08fvP322ygvL8fAgQORn5+Pbdu2oW3btrh06ZI4r9TUVAwYMABLlixBVFQU/v33X0ybNg3Tpk1TOhNd4eeff4aOjg5sbW2V6mpzHD1L+wrl5eXYt28fsrOzoaWlpVD3/9i787ga8/9//I9TORVtWrQQDRUlUbbRjEqjqZAtEZF9GyZbwttEDJKxv21vTOVtMmXGMj4aWaKQQaGyhGqkN7JlqQaVU78/+nV9O9pOKYnH/XZzG+e6Xq/X9bxe1znmPM/rdb2uli1bQldXF2fOnEGbNm3KrZ+Xlyd1+0rpH1iIiKhhefDgQa2Wa8g+h3OszOd+/nWhQSbeqampKCoqQrt27apd94cffhD+bmRkBB8fH4SFhQmJd0ZGBubOnSu0bWJiIpTPyMiAm5sbOnToAABo3bp1hcdZtWoVunTpIpVAlR61fpe5uTm+/PJLBAUFCYn33r17UVRUJHXvZ0Xk5ORgamoqjAKWGD58OOTl5fH27Vvk5eWhX79+WLBgQZXt1YZx48YhKCgII0eOREhICPr06QMdHZ0y5aq6Js2bN8d//vMfeHl54eHDh/jzzz9x5cqVckfxShQUFGDr1q1CsjBkyBDs3r0bjx49goqKCszNzdGrVy+cOnVKSLzHjRsn1G/dujU2btyIrl27Ijc3t9pT5MeNG4c5c+Zg4cKF+P3339GmTRt06tSpTLnSC5UZGRlh2bJlmDJlitT7pqCgAFu2bEHHjh0BADdv3sSJEycQFxeHLl26AAB27twp9V6tiI+PD/r27QsAWLJkCdq3b4/U1FS0a9cOzZs3h4+Pj1D2+++/x9GjR7F3795qJd4vX76EiooKioqK8OrVKwCAt7e38CPZkiVLMH/+fGH2Q+vWrfHjjz/C19e33MQ7PDwchYWF2Llzp3A7QXBwMDQ0NBAdHY1vv/0WHh4e2LNnj5B4R0VF4cWLF3BzcwMAmc/N0tJSiMHExASbNm1CVFQUHB0dceLECVy8eBHJyckwNTUVYi8REBAAT09P4ZqamJhg48aNsLOzw9atW6GkpFTm3O7evQtdXd0y08xLyPo+qkn5efPm4YcffkBeXh7evn0LTU1NTJgwoUw5AwMD3L17t8JjBgQEYMmSJRXuJyKihmP79u31HcJHg31Bta1BJt4lU45rIjw8HBs3bkRaWhpyc3Px9u1bqKmpCftnz56NCRMmYPfu3ejduzfc3d2F5M3b2xtTp07FsWPH0Lt3b7i5ucHS0rLc4yQkJFT7vtZx48Zh1qxZ+Pe//w1VVVUEBQXB3d0dqqqqMtUvKioqc5/zunXr0Lt3b0gkEqSmpmL27NkYNWpUuYswlWjfvr3wRbukr0snnj179ixzL2h5Ro4cifnz5+Pvv/9GSEhIhfeKVnVNAMDd3R0HDhzAypUrsXXr1iqTzMaNG0uN0Onq6sLIyEjqPHR1daVGiS9dugR/f38kJibi+fPnwr3UGRkZMDc3r/J8S+vbty8mT56M06dPIygoSCqpL+3EiRMICAjAzZs3kZ2djbdv3+LNmzd49eoVGjduDKD4HvbS77Nbt25BQUEB1tbWwjZjY2M0bdq0yrhKt6Ovrw+geFS6Xbt2kEgkWLFiBfbu3Yv79+8jPz8feXl5QhyyUlVVxeXLl1FQUIAjR44gNDRUGHkGim/niI2NldomkUjKnHfp8qmpqWU+B2/evBGmQHt6euLLL7/EgwcPYGBggNDQUPTt21eY/SLrub37edbX1xfeIwkJCWjRooWQdL8rMTERSUlJUouRFRUVobCwEHfu3Cl3VPv169flJuQlZH0f1aT83LlzMWbMGGRmZmLu3Ln47rvvYGxsXKacsrKy8ANKeRYsWCAs/ggUj3gbGhpWGicREX2cJk2aBAMDgyrLPXjw4JNPTGXti0/V53CNP7QGmXibmJhAJBLh5s2b1ar3119/wdPTE0uWLIGTkxPU1dURFhaGNWvWCGX8/f0xYsQIRERE4MiRI1i8eDHCwsIwaNAgTJgwAU5OToiIiMCxY8cQEBCANWvW4Pvvvy9zLGVl5Wqfl4eHB2bNmoW9e/fC1tYWsbGxCAgIkKmuRCJBSkoKunbtKrVdT09P+DLdtm1b5OTkYPjw4Vi2bFm5X7IB4M8//xTusb5//z7s7e2RkJBQ7XPT0tJCv379MH78eLx580aYpluaLNcEKJ5ufenSJcjLyyMlJaXKY5fcHlBCJBKVu60kuf7nn3/g5OQEJycnhIaGQkdHBxkZGXByckJ+fr5M51uagoICRo0ahcWLF+PChQs4cOBAmTLp6eno168fpk6diuXLl0NTUxNnz57F+PHjkZ+fLySFysrKtbZwXOk+KGmzpA9++uknbNiwAevXrxfuO585c2a1z19OTk54b5mZmSEtLQ1Tp07F7t27ARTfQrFkyRIMHjy4TN3yktDc3Fx07ty53NW1S2ZQdO3aFW3atEFYWBimTp2KAwcOSK2ZIOu5VfYeqep9n5ubi8mTJ8Pb27vMvpYtW5ZbR1tbu9x1H0rI8j6qaXltbW0YGxvD2NgYv/32Gzp06IAuXbqU+ZHp2bNn5c5UKaGoqFgvi/gREVHtMzAwgJGRUX2H8VFgX1Bta5CJt6amJpycnLB582apKawlXrx4Ue593ufOnUOrVq2wcOFCYVt5UyhNTU1hamqKWbNmYfjw4QgODsagQYMAFK8cPGXKFEyZMgULFizAjh07yk28LS0tERUVVa0pmKqqqnB3d0dQUBDS0tJgamqKnj17ylR3165deP78uTC1tiIl96OWXvTsXa1atRL+XjKdu6IkvSrjxo1Dnz59MG/ePOHYpcl6TebMmQM5OTkcOXIEffr0Qd++fYWFs2rDzZs3kZWVhZUrVwqjdSWL6tXUuHHjsHr1agwbNqzc0ehLly6hsLAQa9asEaYa7927t8p227Zti7dv3+LKlSvo3LkzgOLbLypL4GQRGxuLAQMGYOTIkQCKE/Lbt29Xe7T/XfPnz0ebNm0wa9YsWFtbw9raGrdu3ZL5PWVtbY3w8HA0a9aszEyI0jw9PREaGooWLVpATk5OmFJfW+dmaWmJe/fu4fbt2+WOeltbW+PGjRvV+qxYWVnh4cOHeP78eYUzFqp6H71veaD437Vhw4ZhwYIF+OOPP4TtJbMKrKysZDshIiIiIipXg13VfPPmzZBIJOjWrRv27duHlJQUJCcnY+PGjejRo0e5dUxMTJCRkYGwsDCkpaVh48aNUiNCr1+/xvTp0xEdHY27d+8iNjYWcXFxwhTRmTNn4ujRo7hz5w4uX76MU6dOlTt9FCiefhkXF4fvvvsOSUlJuHnzJrZu3YqnT59Wel7jx4/HuXPnsG3btgqnib569QoPHz7EvXv3cP78ecybNw9TpkzB1KlT0atXL6myL168wMOHD/HgwQPExMRg6dKlMDU1rTDu2ubs7IwnT55g6dKl5e6v6poAQEREBIKCghAaGgpHR0fMnTsXo0ePfu9Es7SWLVtCLBbj3//+N/7++28cOnSozCJw1WVmZoanT59WuLCWsbExCgoKhGPu3r0b27Ztq7Lddu3aoXfv3pg0aRIuXryIK1euYNKkSe89Mm5iYoLjx4/j3LlzSE5OxuTJk/Ho0aMat1fC0NAQgwYNEhbDW7RoEf773/9iyZIluH79OpKTkxEWFiZ1r39pnp6e0NbWxoABA3DmzBncuXMH0dHR8Pb2xr1796TKXb58GcuXL8eQIUOkRmFr49zs7Oxga2sLNzc3HD9+HHfu3MGRI0cQGRkJoPie6XPnzmH69OlISEhASkoK/vjjD0yfPr3CNq2srKCtrY3Y2NgKy1T1Pnrf8iVmzJiB//u//5P6wen8+fNQVFSs8N9UIiIiIpJNg028W7dujcuXL6NXr16YM2cOLCws4OjoiKioKGzdurXcOv3798esWbMwffp0dOrUCefOnYOfn5+wX15eHllZWfDy8oKpqSmGDh0KFxcXYdRaIpFg2rRpMDMzg7OzM0xNTaUWwSrN1NQUx44dQ2JiIrp164YePXrgjz/+qHRBMAD4+uuv0bZtW2RnZ8PLy6vcMjt27IC+vj7atGmDwYMH48aNGwgPDy83lrFjx0JfXx8tWrTA8OHD0b59exw5cqTKOGqLSCSCtrZ2mdWSS1R1TZ48eYLx48fD399fuKd5yZIl0NXVxZQpU2otTh0dHYSEhOC3336Dubk5Vq5cidWrV793u1paWhVOUe7YsSPWrl2LwMBAWFhYIDQ0VOZbC/773/9CV1cXtra2GDRokLASfGX3C1flhx9+gLW1NZycnGBvbw89PT0MHDiwxu2VNmvWLERERODixYtwcnLC4cOHcezYMXTt2hVffvkl1q1bJzXTorTGjRvj9OnTaNmyJQYPHgwzMzPh9oXSI+DGxsbo1q0bkpKS4OnpWSfntm/fPnTt2hXDhw+Hubk5fH19hScVWFpaIiYmBrdv30bPnj1hZWWFRYsWVXp/mLy8PMaOHVvuNPrSKnsf1UZ5oHiBx2+//VbqaQG//vorPD09q32fPxERERFJExW9z0plRPRRuHfvHgwNDXHixAl888039R0OVcPDhw/Rvn17XL58ucIfH+rD06dP0bZtW8THx+OLL76QuV52djbU1dXx8uXLSm8NICKij0d6ejr8/f3h7+8v033NJeXtR2hDo1mjKsu/eFyA6D1PZS5fn0pilbUvPlXVfU98rqrzvadB3uNN9Lk7efIkcnNz0aFDB2RmZsLX1xdGRkbCo+io4dDT08PPP/+MjIyMjyrxTk9Px5YtW6qVdBMRERFR+Zh4EzVABQUF+Ne//oW///4bqqqqsLGxQWhoaJlVualhqK0p/bWpS5cuwnPiiYiIiOj9MPEmaoBKHn1GREREREQfvwa7uBoREREREb0/fX19+Pv7Q19fv75DoY8E3xO1jyPeRERERESfMUVFRS6gRVL4nqh9HPEmIiIiIiIiqkNMvImIiIiIiIjqEBNvIiIiIiIiojrExJuIiIiIiIioDjHxJiIiIiIiIqpDXNWciIiIiIiqLefZ22qVk7V8fWoIMVLDxMSbiIiIiIhkpqqqikbiRrgU+aJa9apbvr40EjeCqqpqfYdBnxhRUVFRUX0HQUREn4bs7Gyoq6vj5cuXUFNTq+9wiIiojmRlZSEnJ6e+w6gTqqqq0NLSqu8wqAGozvcejngTEREREVG1aGlpMTklqgYurkZERERERERUh5h4ExEREREREdUhJt5EREREREREdYiJNxEREREREVEdYuJNREREREREVIe4qjkRERHRZ+hTfhxUQ8PHVxF9+ph4ExEREX1msrKysGDBAuTn59d3KARALBYjICCAyTfRJ4yJNxEREdFnJicnB/n5+RjbZyj0tHTqO5yPysOsJwj+c+8H65uS4+Xk5DDxJvqEMfEmIiIi+kzpaemgpW7z+g7jo8S+IaLaVKPF1Vq3bo2srKwy21+8eIHWrVu/d1BEREREREREn4oaJd7p6emQSCRltufl5eH+/fvvHRQRERERERHRp6JaU80PHTok/P3o0aNQV1cXXkskEkRFRcHIyKjWgiMiIiIiIiJq6KqVeA8cOBAAIBKJMHr0aKl9jRo1gpGREdasWVNrwRERERERERE1dNVKvAsLCwEAX3zxBeLi4qCtrV0nQRERERERERF9Kmp0j/edO3eYdBMRERFVIS8vD+np6cjLy6vvUIg+Kvxs0Oemxo8Ti4qKQlRUFB4/fiyMhJcICgp678CIiIiIGrrMzEz4+/vD39+f6+AQlcLPBn1uapR4L1myBEuXLkWXLl2gr68PkUhU23ERERERERERfRJqlHhv27YNISEhGDVqVG3HQ0RERERERPRJqdE93vn5+bCxsantWKieiUQiHDx4sM6PEx0dDZFIhBcvXtT5sT43ISEh0NDQqLLch7rWVFZ6ejpEIhESEhLqO5QqNaRYiYiIiD5mNUq8J0yYgD179tR2LFSHHj58iO+//x6tW7eGoqIiDA0N4erqiqioqA8ei42NDTIzM6WeAy+rffv2QV5eHvfv3y93v4mJCWbPng0AsLe3h0gkgkgkgqKiIpo3bw5XV1fs37+/TL2SciKRCAoKCmjZsiVmz55d5wt+lBzz/PnzUtvz8vKgpaUFkUiE6OjoGrfv7++PTp06ldmemZkJFxeXSuv+73//w7hx42BgYACxWIxWrVphxowZyMrKqnE8H7uSRLPkj6amJuzs7HDmzJlaO4ahoSEyMzNhYWFRa20SERER0cetRon3mzdvsHbtWtjZ2eH777/H7Nmzpf7QxyU9PR2dO3fGyZMn8dNPP+Hq1auIjIxEr169MG3atA8ej1gshp6eXo3WBujfvz+0tLSwa9euMvtOnz6N1NRUjB8/Xtg2ceJEZGZmIi0tDfv27YO5uTk8PDwwadKkMvWDg4ORmZmJO3fuYMuWLdi9ezeWLVsmc2wlSVt1GRoaIjg4WGrbgQMHoKKiUu22ZKWnpwdFRcUK9//999/o0qULUlJS8OuvvyI1NRXbtm1DVFQUevTogWfPntVZbEDxrJraEB0dXaMFW06cOIHMzEycPn0aBgYG6NevHx49elQrMcnLy0NPTw8KCjVe2xISiaTMopZERERE9PGqUeKdlJSETp06QU5ODteuXcOVK1eEP5yS+PH57rvvIBKJcPHiRbi5ucHU1BTt27fH7Nmzy4y0ljZv3jyYmpqicePGaN26Nfz8/FBQUCDsT0xMRK9evaCqqgo1NTV07twZ8fHxAIC7d+/C1dUVTZs2RZMmTdC+fXv8+eefAMqfah4bGwt7e3s0btwYTZs2hZOTE54/f14mpkaNGmHUqFEICQkpsy8oKAjdu3dH+/bthW2NGzeGnp4eWrRogS+//BKBgYH4z3/+gx07duDEiRNS9TU0NKCnpwdDQ0P069cPAwYMwOXLl2Xq4/cxevRohIWF4fXr11LnMnr0aKly5fVbQkICRCIR0tPTy7QbEhKCJUuWIDExURjBLem3qqaaT5s2DWKxGMeOHYOdnR1atmwJFxcXnDhxAvfv38fChQuFsuW1paGhIXWN/ve//2Ho0KHQ0NCApqYmBgwYIBXzmDFjMHDgQCxfvhwGBgZo27Ytli5dWu6ocKdOneDn51dh7LVBS0sLenp6sLCwwL/+9S9kZ2fjwoULwv5r167BxcUFKioq0NXVxahRo/D06VNhf2FhIVatWgVjY2MoKiqiZcuWWL58OYCy07dLrmtERAQsLS2hpKSEL7/8EteuXRPaK7mF4NChQzA3N4eioiIyMjLw/PlzeHl5oWnTpmjcuDFcXFyQkpIidS6VfbYiIyPx9ddfQ0NDA1paWujXrx/S0tLqqluJiIiIPls1SrxPnTpV4Z+TJ0/Wdoz0Hp49e4bIyEhMmzYNTZo0KbO/svuBVVVVERISghs3bmDDhg3YsWMH1q1bJ+z39PREixYtEBcXh0uXLmH+/Plo1KgRgOLELS8vD6dPn8bVq1cRGBhY4QhuQkICvvnmG5ibm+Ovv/7C2bNn4erqColEUm758ePHIyUlBadPnxa25ebm4vfff5ca7a7I6NGj0bRp03KnnJe4ffs2Tp48ie7du1fZ3vvq3LkzjIyMsG/fPgBARkYGTp8+/d6LFw4bNgxz5sxB+/btkZmZiczMTAwbNqzKes+ePcPRo0fx3XffQVlZWWqfnp4ePD09ER4ejqKiIpniKCgogJOTE1RVVXHmzBnExsZCRUUFzs7OUiPbUVFRuHXrFo4fP47Dhw9j3LhxSE5ORlxcnFDmypUrSEpKwtixY2Xshffz+vVr/Pe//wVQPFMDAF68eAEHBwdYWVkhPj4ekZGRePToEYYOHSrUW7BgAVauXAk/Pz/cuHEDe/bsga6ubqXHmjt3LtasWYO4uDjo6OjA1dVV6oeuV69eITAwEDt37sT169fRrFkzjBkzBvHx8Th06BD++usvFBUVoU+fPkK9qj5b//zzD2bPno34+HhERUVBTk4OgwYN4mg6ERERUS2r+VxHAKmpqUhLS4OtrS2UlZVRVFTER4t9ZFJTU1FUVIR27dpVu+4PP/wg/N3IyAg+Pj4ICwuDr68vgOIEce7cuULbJiYmQvmMjAy4ubmhQ4cOAIDWrVtXeJxVq1ahS5cu2LJli7Ct9Kj1u8zNzfHll18iKCgItra2AIC9e/eiqKgIHh4eVZ6XnJwcTE1Ny4wSDx8+HPLy8nj79i3y8vLQr18/LFiwoMr2asO4ceMQFBSEkSNHIiQkBH369IGOjs57tamsrAwVFRUoKChAT09P5nopKSkoKiqCmZlZufvNzMzw/PlzPHnyBM2aNauyvfDwcBQWFmLnzp3Cvw/BwcHQ0NBAdHQ0vv32WwBAkyZNsHPnTiHBBQAnJycEBweja9euQj07O7tK30+1wcbGBnJycnj16hWKiorQuXNnfPPNNwCATZs2wcrKCitWrBDKBwUFwdDQELdv34a+vj42bNiATZs2CbMW2rRpg6+//rrSYy5evBiOjo4AgF27dqFFixY4cOCAkNAXFBRgy5Yt6NixI4Di63To0CHExsYKi12GhobC0NAQBw8ehLu7e5WfLTc3N6kYgoKCoKOjgxs3bsh8D3peXp7UWgjZ2dky1SP63Dx48KC+Q5DyscVDn981+dzOl6hGiXdWVhaGDh2KU6dOQSQSISUlBa1bt8b48ePRtGlTrFmzprbjpBqSdVSyPOHh4di4cSPS0tKQm5uLt2/fQk1NTdg/e/ZsTJgwAbt370bv3r3h7u6ONm3aAAC8vb0xdepUHDt2DL1794abmxssLS3LPU5CQgLc3d2rFdu4ceMwa9Ys/Pvf/4aqqiqCgoLg7u4OVVVVmeqX9yPRunXr0Lt3b0gkEqSmpmL27NkYNWoUwsLCKmynffv2uHv3rtAmAKmR/Z49e+LIkSNVxjNy5EjMnz8ff//9N0JCQrBx40aZzqMuVfXeKZ0gVyYxMRGpqallrs2bN2+kpjV36NChTJsTJ07EuHHjsHbtWsjJyWHPnj1Ssy7KU7r/JRIJ8vLypLaNHDkS27Ztq7SN8PBwtGvXDteuXYOvry9CQkKE2RyJiYk4depUuTM40tLS8OLFC+Tl5QmJuqx69Ogh/F1TUxNt27ZFcnKysE0sFkt9hpKTk6GgoCA1K0NLS0uqXlWfrZSUFCxatAgXLlzA06dPhZHujIwMmRPvgIAALFmyRLaTJPqMbd++vb5DoI8c3yNEn7YaJd6zZs1Co0aNkJGRITUqNmzYMMyePZuJ90fExMQEIpEIN2/erFa9v/76C56enliyZAmcnJygrq6OsLAwqWvr7++PESNGICIiAkeOHMHixYsRFhaGQYMGYcKECXByckJERASOHTuGgIAArFmzBt9//32ZY707nVkWHh4emDVrFvbu3QtbW1vExsYiICBAproSiQQpKSnCKGoJPT09GBsbAwDatm2LnJwcDB8+HMuWLRO2v+vPP/8UpvXev38f9vb2UuscyHpuJffXjh8/Hm/evIGLiwtycnKkysjJFd8ZUjohLj0VubYYGxtDJBIhOTkZgwYNKrM/OTkZOjo6wm0KIpGoTJJeOq7c3Fx07twZoaGhZdoqPapf3q0Qrq6uUFRUxIEDByAWi1FQUIAhQ4ZUGn/p/r9w4QLmzZsntTJ86R+PKmJoaAgTExOYmJjg7du3GDRoEK5duwZFRUXk5ubC1dUVgYGBZerp6+vj77//rrL9mlBWVq72jKKq3n+urq5o1aoVduzYAQMDAxQWFsLCwqJai9stWLBAalHN7OxsGBoaVitOos/BpEmTYGBgUN9hCB48eMBE7yPzsb1H6hrfg/S5qVHifezYMRw9ehQtWrSQ2m5iYiKM/tHHQVNTE05OTti8eTO8vb3LJDcvXrwo9z7vc+fOoVWrVlKLaJV3bU1NTWFqaopZs2Zh+PDhCA4OFpI1Q0NDTJkyBVOmTMGCBQuwY8eOchNvS0tLREVFVWvUTFVVFe7u7ggKCkJaWhpMTU3Rs2dPmeru2rULz58/LzPN9l3y8vIAILXo2btatWol/L1kleqKkvSqjBs3Dn369MG8efOEY5dWkqRmZmaiadOmAFDlYoZisbjCe+UroqWlBUdHR2zZsgWzZs2SSt4ePnyI0NBQqdXwdXR0kJmZKbxOSUnBq1evhNfW1tYIDw9Hs2bNZEp6S1NQUMDo0aMRHBwMsVgMDw+PKpPJ0v1/7949KCgo1PiaAMCQIUOwaNEioT+sra2xb98+GBkZlbsyuYmJCZSVlREVFYUJEybIfJzz58+jZcuWAIDnz5/j9u3bFU73B4qn/L99+xYXLlwQpppnZWXh1q1bMDc3B1D5Z6uk7I4dO4TPztmzZ2WOt4SiomKlK+QTUTEDA4MaPWWBPh98jxB92mq0uNo///yDxo0bl9n+7NkzfgH7CG3evBkSiQTdunXDvn37kJKSguTkZGzcuFFqemtpJiYmyMjIQFhYGNLS0rBx40YcOHBA2P/69WtMnz4d0dHRuHv3LmJjYxEXFyckCjNnzsTRo0dx584dXL58GadOnaowiViwYAHi4uLw3XffISkpCTdv3sTWrVulVokuz/jx43Hu3Dls27YN48aNK7fMq1ev8PDhQ9y7dw/nz5/HvHnzMGXKFEydOhW9evWSKvvixQs8fPgQDx48QExMDJYuXQpTU9NKk5/a5OzsjCdPnmDp0qXl7jc2NoahoSH8/f2RkpKCiIiIKmeXGBkZ4c6dO0hISMDTp09lfi75pk2bkJeXBycnJ5w+fRr/+9//EBkZCUdHR5iammLRokVCWQcHB2zatAlXrlxBfHw8pkyZIkzLBooX4dPW1saAAQNw5swZ3LlzB9HR0fD29sa9e/eqjGXChAk4efIkIiMjK7zOdUkkEsHb2xsrV67Eq1evMG3aNDx79gzDhw9HXFwc0tLScPToUYwdOxYSiQRKSkqYN28efH198d///hdpaWk4f/48fv7550qPs3TpUkRFReHatWsYM2YMtLW1MXDgwArLm5iYYMCAAZg4cSLOnj2LxMREjBw5Es2bN8eAAQMAVP7Zatq0KbS0tLB9+3akpqbi5MmTfBwkERERUR2pUeLds2dPYaVfoPiLacnjc95NZqj+tW7dGpcvX0avXr0wZ84cWFhYwNHREVFRUdi6dWu5dfr3749Zs2Zh+vTp6NSpE86dOyf1CCd5eXlkZWXBy8sLpqamGDp0KFxcXISRNYlEgmnTpsHMzAzOzs4wNTWVWuCpNFNTUxw7dgyJiYno1q0bevTogT/++KPK5xx//fXXaNu2LbKzs+Hl5VVumR07dkBfXx9t2rTB4MGDcePGDYSHh5cby9ixY6Gvr48WLVpg+PDhaN++PY4cOfJez1uuDpFIBG1t7QrvnW7UqBF+/fVX3Lx5E5aWlggMDKzyOeNubm5wdnZGr169oKOjg19//VWmWExMTBAXF4fWrVtj6NChaNWqFVxcXGBqaiqsSl5izZo1MDQ0RM+ePTFixAj4+PhI/TDXuHFjnD59Gi1btsTgwYNhZmYmTKmXZQTcxMQENjY2aNeu3QdZZb48o0ePRkFBATZt2gQDAwPExsZCIpHg22+/RYcOHTBz5kxoaGgItwP4+flhzpw5WLRoEczMzDBs2DA8fvy40mOsXLkSM2bMQOfOnfHw4UP83//9X5X30QcHB6Nz587o168fevTogaKiIvz555/CDx+Vfbbk5OQQFhaGS5cuwcLCArNmzcJPP/1UOx1GRERERFJERTVYfevatWv45ptvYG1tjZMnT6J///64fv06nj17htjYWGGBLSL6dCxevBhr167F8ePH8eWXX36w4xYVFcHExATffffdJzkiGx0djV69euH58+eVPt6vocjOzoa6ujpevnxZ7VsLiD5F6enp8Pf3h7+//0c1jbgkrgWjpqGlbvP6DuejkvHoPgJ2b/5gfVNyvI/tPVLXPtbPBlF1VOd7T42G8iwsLHD79m1s2rQJqqqqyM3NxeDBgzFt2jTo6+vXKGgi+rgtWbIERkZGOH/+PLp16yaM7talJ0+eICwsDA8fPvxgz+4mIiIiIqptNZ5Dq66uLrXwFhF9+j508tusWTNoa2tj+/btwoJyREREREQNTY0T7zdv3iApKQmPHz8Wnv1aon///u8dGBHR+zyHvqGwt7f/LM6TiIiI6HNWo8Q7MjISXl5e5a46LRKJqv34IiIiIqJPkb6+Pvz9/XkrHtE7+Nmgz02NbtL8/vvv4e7ujszMTBQWFkr9YdJNREREVExRURFGRkZ83CrRO/jZoM9NjRLvR48eYfbs2dDV1a3teIiIiIiIiIg+KTVKvIcMGYLo6OhaDoWIiIiIiIjo01Oje7w3bdoEd3d3nDlzBh06dECjRo2k9nt7e9dKcEREREREREQNXY0S719//RXHjh2DkpISoqOjIRKJhH0ikYiJNxEREREREdH/r0aJ98KFC7FkyRLMnz8fcnI1mq1ORERERERE9FmoUeKdn5+PYcOGMekmIiIiasAeZj2p7xA+OiV98qH6hteA6PNQo8R79OjRCA8Px7/+9a/ajoeIiIiI6piqqirEYjGC/9xb36F8tD5k34jFYqiqqn6w4xHRh1ejxFsikWDVqlU4evQoLC0tyyyutnbt2loJjoiIiIhqn5aWFgICApCTk1PfoRCKfwjR0tKq7zCIqA7VKPG+evUqrKysAADXrl2T2ld6oTUiIiIi+jhpaWkx2SMi+kBqlHifOnWqtuMgIiIiIiIi+iRxdTQiIiIiIiKiOiTziPfgwYMREhICNTU1DB48uNKy+/fvf+/AiIiIiIiIiD4FMife6urqwv3b6urqdRYQERERERER0adEVFRUVFSdCkVFRfjf//4HHR0dKCsr11VcRETUAGVnZ0NdXR0vX76EmppafYdDREREVGeq872n2ourFRUVwdjYGNevX4eJiUmNgyQiIiIiqm9ZWVkf5LFqfGQY0eet2om3nJwcTExMkJWVxcSbiIiIiBqsrKwsLFiwAPn5+XV+LLFYjICAACbfRJ+pGj1ObOXKlZg7dy62bt0KCwuL2o6JiIiIiKjO5eTkID8/H/3794e2trbM9Z4+fYpDhw7JXK+kfE5ODhNvos9UjRJvLy8vvHr1Ch07doRYLC5zr/ezZ89qJTgiIiIiorqmra0NPT29D1aPiD4/NUq8169fX8thEBEREREREX2aapR4jx49urbjICIiIiIiIvokydW0YlpaGn744QcMHz4cjx8/BgAcOXIE169fr7XgiIiIiIiIiBq6GiXeMTEx6NChAy5cuID9+/cjNzcXAJCYmIjFixfXaoBEREREREREDVmNEu/58+dj2bJlOH78OMRisbDdwcEB58+fr7XgiIiIiIiIiBq6GiXeV69exaBBg8psb9asGZ4+ffreQRERERERvY+8vDykp6cjLy+vvkOpV+wHoo9DjRJvDQ0NZGZmltl+5coVNG/e/L2DIiIiIiJ6H5mZmfD39y/3O+vnhP1A9HGoUeLt4eGBefPm4eHDhxCJRCgsLERsbCx8fHzg5eVV2zESERERERERNVg1SrxXrFiBdu3awdDQELm5uTA3N4etrS1sbGzwww8/1HaMRERERERERA1WjRJvsViMHTt2IC0tDYcPH8Yvv/yCmzdvYvfu3ZCXl6/tGD9rIpEIBw8erPPjREdHQyQS4cWLF3V+rM9NSEgINDQ0qiz3oa7154x9XDn+O0BERERUN2r8HG8AaNmyJVxcXODu7g4TE5Paiumz8fDhQ3z//fdo3bo1FBUVYWhoCFdXV0RFRX3wWGxsbJCZmQl1dfVq1923bx/k5eVx//79cvebmJhg9uzZAAB7e3uIRCKIRCIoKiqiefPmcHV1xf79+8vUKyknEomgoKCAli1bYvbs2XW+OEjJMd9doT8vLw9aWloQiUSIjo6ucfv+/v7o1KlTme2ZmZlwcXGptO7//vc/jBs3DgYGBhCLxWjVqhVmzJiBrKysGsfTUOzbtw/29vZQV1eHiooKLC0tsXTpUjx79qy+QyMiIqJaVlhYiOTkZJw/fx7JyckoLCys75CI3kuNE++ff/4ZFhYWUFJSgpKSEiwsLLBz587ajO2Tlp6ejs6dO+PkyZP46aefcPXqVURGRqJXr16YNm3aB49HLBZDT08PIpGo2nX79+8PLS0t7Nq1q8y+06dPIzU1FePHjxe2TZw4EZmZmUhLS8O+fftgbm4ODw8PTJo0qUz94OBgZGZm4s6dO9iyZQt2796NZcuWyRxbenp6jc7J0NAQwcHBUtsOHDgAFRWVarclKz09PSgqKla4/++//0aXLl2QkpKCX3/9Fampqdi2bRuioqLQo0ePOk9A8/Pza6Wd6OhoGBkZVavOwoULMWzYMHTt2hVHjhzBtWvXsGbNGiQmJmL37t21EldN1Va/SCSSGn2pqGk9IiKij1V8fDx8fX0RGBiIbdu2ITAwEL6+voiPj6/v0IhqrEaJ96JFizBjxgy4urrit99+w2+//QZXV1fMmjULixYtqu0YP0nfffcdRCIRLl68CDc3N5iamqJ9+/aYPXt2pc9CnzdvHkxNTdG4cWO0bt0afn5+KCgoEPYnJiaiV69eUFVVhZqaGjp37iz8I3X37l24urqiadOmaNKkCdq3b48///wTQPlTTGNjY2Fvb4/GjRujadOmcHJywvPnz8vE1KhRI4waNQohISFl9gUFBaF79+5o3769sK1x48bQ09NDixYt8OWXXyIwMBD/+c9/sGPHDpw4cUKqvoaGBvT09GBoaIh+/fphwIABuHz5skx9/D5Gjx6NsLAwvH79WupcRo8eLVWuvH5LSEiASCRCenp6mXZDQkKwZMkSJCYmCiPrJf1W1TToadOmQSwW49ixY7CzsxNmnJw4cQL379/HwoULhbLltaWhoSF1jf73v/9h6NCh0NDQgKamJgYMGCAV85gxYzBw4EAsX74cBgYGaNu2LZYuXQoLC4sysXXq1Al+fn4Vxv4+Ll68iBUrVmDNmjX46aefYGNjAyMjIzg6OmLfvn1S12Tr1q1o06YNxGIx2rZtW25S/vTpUwwaNAiNGzeGiYkJDh06JLX/2rVrcHFxgYqKCnR1dTFq1CipxyTa29tj+vTpmDlzJrS1teHk5CRTvXeV3IJw6NAhmJubQ1FRERkZGcjLy4OPjw+aN2+OJk2aoHv37lIzLCqqFxcXB0dHR2hra0NdXR12dnZlPisikQg7d+6s9PxLe/XqFVxcXPDVV19x+jkREX0Q8fHx2Lx5M1q0aIEffvgB27Ztww8//IAWLVpg8+bNTL6pwapR4r1161bs2LEDAQEB6N+/P/r374+AgABs374dW7Zsqe0YPznPnj1DZGQkpk2bhiZNmpTZX9n9wKqqqggJCcGNGzewYcMG7NixA+vWrRP2e3p6okWLFoiLi8OlS5cwf/58NGrUCEBx4paXl4fTp0/j6tWrCAwMrHAENyEhAd988w3Mzc3x119/4ezZs3B1dYVEIim3/Pjx45GSkoLTp08L23Jzc/H7779LjXZXZPTo0WjatGm5U85L3L59GydPnkT37t2rbO99de7cGUZGRti3bx8AICMjA6dPn8aoUaPeq91hw4Zhzpw5aN++PTIzM5GZmYlhw4ZVWe/Zs2c4evQovvvuOygrK0vt09PTg6enJ8LDw1FUVCRTHAUFBXBycoKqqirOnDmD2NhYqKiowNnZWWoENyoqCrdu3cLx48dx+PBhjBs3DsnJyYiLixPKXLlyBUlJSRg7dqyMvVA9oaGhUFFRwXfffVfu/pLPy4EDBzBjxgzMmTMH165dw+TJkzF27FicOnVKqvySJUswdOhQJCUloU+fPvD09BRmC7x48QIODg6wsrJCfHw8IiMj8ejRIwwdOlSqjV27dkEsFiM2Nhbbtm2Tud67Xr16hcDAQOzcuRPXr19Hs2bNMH36dPz1118ICwtDUlIS3N3d4ezsjJSUlErr5eTkYPTo0Th79izOnz8PExMT9OnTBzk5OTKff2kvXryAo6MjCgsLcfz4cZnWKSAiInofhYWFCAsLQ8eOHeHt7Q1jY2MoKSnB2NgY3t7e6NixI8LDwznTixokhZpUKigoQJcuXcps79y5M96+ffveQX3qUlNTUVRUhHbt2lW7bulV442MjODj44OwsDD4+voCKE4Q586dK7Rd+t77jIwMuLm5oUOHDgCA1q1bV3icVatWoUuXLlI/pJQetX6Xubk5vvzySwQFBcHW1hYAsHfvXhQVFcHDw6PK85KTk4OpqWmZUeLhw4dDXl4eb9++RV5eHvr164cFCxZU2V5tGDduHIKCgjBy5EiEhISgT58+0NHRea82lZWVoaKiAgUFBejp6clcLyUlBUVFRTAzMyt3v5mZGZ4/f44nT56gWbNmVbZX8j+tnTt3ClPxg4ODoaGhgejoaHz77bcAgCZNmmDnzp0Qi8VCXScnJwQHB6Nr165CPTs7u0rfT+8jJSUFrVu3Fn5Aqsjq1asxZswYIUEvmT2yevVq9OrVSyg3ZswYDB8+HEDxExo2btyIixcvwtnZGZs2bYKVlRVWrFghlA8KCoKhoSFu374NU1NTAMWfq1WrVgllli1bJlO9dxUUFGDLli3o2LEjgOLPaHBwMDIyMmBgYAAA8PHxQWRkJIKDg4X2360HAA4ODlJtb9++HRoaGoiJiUG/fv1kOv8SDx8+xLBhw2BiYoI9e/ZIXf935eXlSa27kJ2dXWFZIqL68ODBgxrtqwsf+nj1dcyaunXrFp4+fYopU6ZATk56fFBOTg79+vXDsmXLcOvWrQq/ExF9rGqUeI8aNQpbt27F2rVrpbZv374dnp6etRLYp0zWUcnyhIeHY+PGjUhLS0Nubi7evn0LNTU1Yf/s2bMxYcIE7N69G71794a7uzvatGkDAPD29sbUqVNx7Ngx9O7dG25ubrC0tCz3OAkJCXB3d69WbOPGjcOsWbPw73//G6qqqggKCoK7uztUVVVlql9UVFTmfux169ahd+/ekEgkSE1NxezZszFq1CiEhYVV2E779u1x9+5doU0AUiP7PXv2xJEjR6qMZ+TIkZg/fz7+/vtvhISEYOPGjTKdR12q6r1TWYJUWmJiIlJTU8tcmzdv3iAtLU143aFDhzJtTpw4EePGjcPatWshJyeHPXv2SM26KE/p/pdIJMjLy5PaNnLkSGzbtq3curJ+XpKTk8usE/DVV19hw4YNUttKv+ebNGkCNTU1PH78GEBxv5w6darcmSBpaWlCAt25c2epfbLWe5dYLJaK5+rVq5BIJGXKlyzsV1E9AHj06BF++OEHREdH4/Hjx5BIJHj16hUyMjJkPv8Sjo6O6NatG8LDw6t8UkVAQACWLFlSaRkiovq0ffv2+g5B8DHF8jF6+fIlAKBFixbl7m/evLlUOaKGpEaJN1C8uNqxY8fw5ZdfAgAuXLiAjIwMeHl5CStYAyiTnFPxaJlIJMLNmzerVe+vv/6Cp6cnlixZAicnJ6irqyMsLAxr1qwRyvj7+2PEiBGIiIjAkSNHsHjxYoSFhWHQoEGYMGECnJycEBERgWPHjiEgIABr1qzB999/X+ZY705nloWHhwdmzZqFvXv3wtbWFrGxsQgICJCprkQiQUpKijCKWkJPTw/GxsYAgLZt2yInJwfDhw/HsmXLhO3v+vPPP4X73u/fvw97e3skJCRU+9y0tLTQr18/jB8/Hm/evIGLi0uZabslv8aWTg5L33NfW4yNjSESiZCcnIxBgwaV2Z+cnAwdHR1hOrBIJCqTsJaOKzc3F507d0ZoaGiZtkqP6pd3K4SrqysUFRVx4MABiMViFBQUYMiQIZXGX7r/L1y4gHnz5kndt1z6x6N3mZqa4uzZsygoKKhy1FsW77YhEomEKWu5ublwdXVFYGBgmXr6+vrC39/tF1nrvUtZWVnqx6bc3FzIy8vj0qVLZRLe0kn9u/WA4ts1srKysGHDBrRq1QqKioro0aNHmcXfKjv/En379sW+fftw48YNYYZMRRYsWCD1b352djYMDQ0rrUNE9CFNmjRJmEX0rgcPHnzQZLiyWOrKhz7H91HydJ179+6V+z2v5Ak6NXkKD1F9q1Hife3aNVhbWwOAMDqmra0NbW1tXLt2TShXk9WkPweamppwcnLC5s2b4e3tXeZL/IsXL8q9n/LcuXNo1aqV1CJaJSO7pZmamsLU1BSzZs3C8OHDERwcLCRrhoaGmDJlCqZMmYIFCxZgx44d5SbelpaWiIqKqtZIlqqqKtzd3REUFCSM8vXs2VOmurt27cLz58/h5uZWabmSZKT0omfvatWqlfB3BYXit3hFSXpVxo0bhz59+mDevHnljvyVJKmZmZlo2rQpAOkkszxisbjCe+UroqWlBUdHR2zZsgWzZs2S+vHg4cOHCA0NlVoNX0dHB5mZmcLrlJQUvHr1SnhtbW2N8PBwNGvWrNKktzwKCgoYPXo0goODIRaL4eHhUeWPGaX7/969e1BQUJD5mowYMQIbN27Eli1bMGPGjDL7Sz4vZmZmiI2NlVpsLTY2Fubm5jKeWXG/7Nu3D0ZGRsJ7py7rvcvKygoSiQSPHz+W+bNTIjY2Flu2bEGfPn0AFC+eV9nibpVZuXIlVFRU8M033yA6OrrSPlRUVKx0NX4iovpmYGBQ7adp1JWPKZaPUdu2baGtrY3Dhw/D29tbarp5YWEhDh8+DB0dHbRt27YeoySqmRp9Q3x3sSKqvs2bN+Orr75Ct27dsHTpUlhaWuLt27c4fvw4tm7diuTk5DJ1TExMkJGRgbCwMHTt2hURERE4cOCAsP/169eYO3cuhgwZgi+++AL37t1DXFyckMzOnDkTLi4uMDU1xfPnz3Hq1KkK749ZsGABOnTogO+++w5TpkyBWCzGqVOn4O7uDm1t7QrPa/z48ejZsyeSk5Mxb968csu8evUKDx8+xNu3b3Hv3j0cOHAA69atw9SpU6XuxQWKk6qHDx+isLAQKSkpWLp0KUxNTT/YfT3Ozs548uRJhcmpsbExDA0N4e/vj+XLl+P27dtSMxDKY2RkhDt37iAhIQEtWrSAqqqqTInLpk2bYGNjAycnJyxbtgxffPEFrl+/jrlz58LU1FTqiQIODg7YtGkTevToAYlEgnnz5kmNdHp6euKnn37CgAEDsHTpUrRo0QJ3797F/v374evrW+EUrxITJkwQrkFsbGyVsb+P7t27w9fXF3PmzMH9+/cxaNAgGBgYCI9T+/rrrzFjxgzMnTsXQ4cOhZWVFXr37o3/+7//w/79+8uslF+ZadOmYceOHRg+fDh8fX2hqamJ1NRUhIWFYefOnRVOu65pvXeZmprC09MTXl5eWLNmDaysrPDkyRNERUXB0tISffv2rbCuiYkJdu/ejS5duiA7Oxtz586t0cyVEqtXr4ZEIoGDgwOio6NrtCYFERFRdcjJycHDwwObN2/Gxo0b0a9fPzRv3hz379/H4cOHkZiYiGnTppW5/5uoIajRu/bJkycV7rt69WqNg/mctG7dGpcvX0avXr0wZ84cWFhYwNHREVFRUdi6dWu5dfr3749Zs2Zh+vTp6NSpE86dOyf1CCd5eXlkZWXBy8sLpqamGDp0KFxcXIRRa4lEgmnTpsHMzAzOzs4wNTWtcBV6U1NTHDt2DImJiejWrRt69OiBP/74o8rRvK+//hpt27ZFdnY2vLy8yi2zY8cO6Ovro02bNhg8eDBu3LiB8PDwcmMZO3Ys9PX10aJFCwwfPhzt27fHkSNH3mtUsTpEIhG0tbUrvHe6UaNG+PXXX3Hz5k1YWloiMDCwyueMu7m5wdnZGb169YKOjg5+/fVXmWIxMTFBXFwcWrdujaFDh6JVq1bCDyklq5KXWLNmDQwNDdGzZ0+MGDECPj4+aNy4sbC/cePGOH36NFq2bInBgwfDzMxMmFIvywi4iYkJbGxs0K5duw+yynxgYCD27NmDCxcuwMnJSXj0nqWlpTDCPXDgQGzYsAGrV69G+/bt8Z///AfBwcGwt7eX+TgGBgaIjY2FRCLBt99+iw4dOmDmzJnQ0NCo9H/yNa1XnuDgYHh5eWHOnDlo27YtBg4ciLi4OLRs2bLSej///DOeP38Oa2trjBo1Ct7e3jIttFeZdevWYejQoXBwcMDt27ffqy0iIiJZdOnSBdOmTcO9e/ewbNkyTJ06FcuWLcP9+/cxbdq0chd4JmoIREU1WOlLT08PP//8c5nRl9WrV8PPz6/SacBEVHsWL16MtWvX4vjx48J6Cx9CUVERTExM8N1330nd30uUnZ0NdXV1vHz5stq3MRAR1ab09HT4+/vD39+/wundJWXGjRtXraeNPHz4EEFBQTLXKylfWSx1RZZ++BgVFhbi1q1bePnyJdTV1dG2bVuOdNNHpzrfe2o0bDh79my4ublh7NixWLt2LZ49ewYvLy9cvXoVe/bsqVHQRFR9S5YsgZGREc6fP49u3bp9kP8hPXnyBGFhYXj48GGdPbubiIiIPm9ycnJ8ZBh9UmqUePv6+sLR0RGjRo2CpaUlnj17hu7duyMpKalavxYS0fv70Mlvs2bNoK2tje3btwsLyhERERERUcVqfKOssbExLCwssG/fPgDAsGHDmHQTfQbe5zn0RERERESfoxrNS42NjYWlpSVSUlKQlJSErVu34vvvv8ewYcPw/Pnz2o6RiIiIiKha9PX14e/vD319/foOpV6xH4g+DjVKvB0cHDBs2DCcP38eZmZmmDBhAq5cuYKMjAx06NChtmMkIiIiIqoWRUVFGBkZyfTIzk8Z+4Ho41CjqebHjh2DnZ2d1LY2bdogNjYWy5cvr5XAiIiIiIiIiD4F1Rrx7tOnD16+fCkk3StXrsSLFy+E/c+fP5f5mcREREREREREn4NqJd5Hjx5FXl6e8HrFihV49uyZ8Prt27e4detW7UVHRERERERE1MBVK/F+dzVjrm5MREREREREVLkaLa5GRERERERERLKp1uJqIpEIIpGozDYiIiIioobq6dOnNSova73qtk9En55qJd5FRUUYM2aM8DiCN2/eYMqUKWjSpAkASN3/TURERET0MVNVVYVYLMahQ4dqVL869cRiMVRVVWt0HCJq+ERF1bhRe+zYsTKVCw4OrnFARETUcGVnZ0NdXR0vX76EmppafYdDRFSlrKws5OTk1PlxVFVVoaWlVefHIaIPpzrfe6qVeBMREVWGiTcRERF9LqrzvYeLqxERERERERHVISbeRERERERERHWIiTcRERERERFRHWLiTURERERERFSHmHgTERERERER1aFqPcebiIiIqL58qMc+vQ8+MoqIiMrDxJuIiIg+ellZWVgwfwHyC/LrO5RKiRuJEbAygMk3ERFJYeJNREREH72cnBzkF+Sju0F3qIlr9oz47LxsXMi8gO763aGmWPvPmc/Oz8aFBxeQk5PDxJuIiKQw8SYiIqIGQ02shqbKTd+vDcX3b4OIiKg6uLgaERERERERUR1i4k1ERERERERUh5h4ExEREREREdUhJt5EREREREREdYiJNxEREREREVEdYuJNREREMsvLy0N6ejry8vLqO5TPDvueiKjhYuJNREREMsvMzIS/vz8yMzPrO5TPDvueiKjhYuJNREREREREVIeYeBMRERERERHVISbeRPTRSE9Ph0gkQkJCwnu14+/vj06dOtVKTB+L2uqb6jIyMsL69es/6DGJiIiIPjVMvIlq0ZMnTzB16lS0bNkSioqK0NPTg5OTE2JjY2v1OPb29pg5c2attvkxMDQ0RGZmJiwsLOo7lA8iJCQEGhoaMpX93PqGiIiI6FOiUN8BEH1K3NzckJ+fj127dqF169Z49OgRoqKikJWVVd+hNQjy8vLQ09Or7zDem0QigUgkgpxc7fy2mZ+fD7FY/En0DREREdHniCPeRLXkxYsXOHPmDAIDA9GrVy+0atUK3bp1w4IFC9C/f3+pchMmTICOjg7U1NTg4OCAxMREYX/JNOndu3fDyMgI6urq8PDwQE5ODgBgzJgxiImJwYYNGyASiSASiZCeng4AuHbtGlxcXKCiogJdXV2MGjUKT58+Fdq2t7eHt7c3fH19oampCT09Pfj7+5c5j8mTJ0NXVxdKSkqwsLDA4cOHhf1nz55Fz549oaysDENDQ3h7e+Off/4pt09evnwJeXl5xMfHAwAKCwuhqamJL7/8Uijzyy+/wNDQEEDZ6dTR0dEQiUSIiopCly5d0LhxY9jY2ODWrVtSx1m5ciV0dXWhqqqK8ePH482bN1L7CwsLsXTpUrRo0QKKioro1KkTIiMjhf1DhgzB9OnThdczZ86ESCTCzZs3ARQnvk2aNMGJEyfKPc+SketDhw7B3NwcioqKyMjIQF5eHnx8fNC8eXM0adIE3bt3R3R0tHBuY8eOxcuXL4XrWHItjIyM8OOPP8LLywtqamqYNGlSuVPNK7ve27dvh4GBAQoLC6ViHTBgAMaNGwcASEtLw4ABA6CrqwsVFRV07dq1wnMkIiIioppj4k1US1RUVKCiooKDBw9W+oxVd3d3PH78GEeOHMGlS5dgbW2Nb775Bs+ePRPKpKWl4eDBgzh8+DAOHz6MmJgYrFy5EgCwYcMG9OjRAxMnTkRmZiYyMzNhaGiIFy9ewMHBAVZWVoiPj0dkZCQePXqEoUOHSh1/165daNKkCS5cuIBVq1Zh6dKlOH78OIDiBNXFxQWxsbH45ZdfcOPGDaxcuRLy8vJCXM7OznBzc0NSUhLCw8Nx9uxZqaS1NHV1dXTq1ElINq9evQqRSIQrV64gNzcXABATEwM7O7tK+3bhwoVYs2YN4uPjoaCgICSOALB37174+/tjxYoViI+Ph76+PrZs2SJVf8OGDVizZg1Wr16NpKQkODk5oX///khJSQEA2NnZCTGWxKStrS1si4uLQ0FBAWxsbCqM8dWrVwgMDMTOnTtx/fp1NGvWDNOnT8dff/2FsLAwJCUlwd3dHc7OzkhJSYGNjQ3Wr18PNTU14Tr6+PgI7a1evRodO3bElStX4OfnV+Z4VV1vd3d3ZGVl4dSpU0KdZ8+eITIyEp6engCA3Nxc9OnTB1FRUbhy5QqcnZ3h6uqKjIyMSq8HEREREVUPp5oT1RIFBQWEhIRg4sSJ2LZtG6ytrWFnZwcPDw9YWloCKB4tvnjxIh4/fgxFRUUAxQnWwYMH8fvvv2PSpEkAihPgkJAQqKqqAgBGjRqFqKgoLF++HOrq6hCLxWjcuLHU1ONNmzbBysoKK1asELYFBQXB0NAQt2/fhqmpKQDA0tISixcvBgCYmJhg06ZNiIqKgqOjI06cOIGLFy8iOTlZKN+6dWuhvYCAAHh6egr3l5uYmGDjxo2ws7PD1q1boaSkVKZf7O3tER0dDR8fH0RHR8PR0RE3b97E2bNn4ezsjOjoaPj6+lbat8uXLxeS8/nz56Nv37548+YNlJSUsH79eowfPx7jx48HACxbtgwnTpyQGvVevXo15s2bBw8PDwBAYGAgTp06hfXr12Pz5s2wt7fHjBkz8OTJEygoKODGjRvw8/NDdHQ0pkyZgujoaHTt2hWNGzeuMMaCggJs2bIFHTt2BABkZGQgODgYGRkZMDAwAAD4+PggMjISwcHBWLFiBdTV1SESicqdQu7g4IA5c+YIr0tmNZSQ5Xq7uLhgz549+OabbwAAv//+O7S1tdGrVy8AQMeOHYV4AeDHH3/EgQMHcOjQoQp/THlXXl6e1A9N2dnZMtWjhu/Bgwef9PHeR13F2pD6gIiIpDHxJqpFbm5u6Nu3L86cOYPz58/jyJEjWLVqFXbu3IkxY8YgMTERubm50NLSkqr3+vVrpKWlCa+NjIyEpBsA9PX18fjx40qPnZiYiFOnTkFFRaXMvrS0NKnEu7TSbSckJKBFixZC2fKOkZSUhNDQUGFbUVERCgsLcefOHZiZmZWpY2dnh59//hkSiQQxMTH49ttvoaenh+joaFhaWiI1NRX29vaVnlvpmPX19QEAjx8/RsuWLZGcnIwpU6ZIle/Ro4cw0pudnY0HDx7gq6++kirz1VdfCVP8LSwsoKmpiZiYGIjFYlhZWaFfv37YvHkzgOIR8KpiFIvFUnFevXoVEomkTF/m5eWVuf7l6dKlS6X7Zbnenp6emDhxIrZs2QJFRUWEhobCw8NDuPc8NzcX/v7+iIiIQGZmJt6+fYvXr19Xa8Q7ICAAS5Yskbk8fTq2b99e3yF8tNg3RET0LibeRLVMSUkJjo6OcHR0hJ+fHyZMmIDFixdjzJgxyM3Nhb6+vtS05hKlV7du1KiR1D6RSFTmXt135ebmwtXVFYGBgWX2lSSrVbWtrKxc5TEmT54Mb2/vMvtatmxZbh1bW1vk5OTg8uXLOH36NFasWAE9PT2sXLkSHTt2hIGBAUxMTCo9bumYRSIRAFTZH9UhEolga2uL6OhoKCoqwt7eHpaWlsjLy8O1a9dw7tw5qWng5VFWVhZiA4r7Sl5eHpcuXRKm6pcoL1l+V5MmTSrdL8v1dnV1RVFRESIiItC1a1ecOXMG69atE8r5+Pjg+PHjWL16NYyNjaGsrIwhQ4YgPz+/yvhKLFiwALNnzxZeZ2dnC/fs06dt0qRJwmyOD+HBgwcNJqGtq75pSH1ARETSmHgT1TFzc3McPHgQAGBtbY2HDx9CQUEBRkZGNW5TLBZDIpFIbbO2tsa+fftgZGQEBYWafbQtLS1x7949qanp7x7jxo0bMDY2lrlNDQ0NWFpaYtOmTWjUqBHatWuHZs2aYdiwYTh8+HCV93dXxczMDBcuXICXl5ew7fz588Lf1dTUYGBggNjYWKljxcbGolu3bsJrOzs77NixA4qKili+fDnk5ORga2uLn376CXl5eWVGzKtiZWUFiUSCx48fo2fPnuWWKe86ykqW662kpITBgwcjNDQUqampaNu2LaytrYX9sbGxGDNmDAYNGgSgOJl/d0p7VRQVFYXbJujzYmBg8F7/jn3K2DdERPQuLq5GVEuysrLg4OCAX375BUlJSbhz5w5+++03rFq1CgMGDAAA9O7dGz169MDAgQNx7NgxpKen49y5c1i4cKGw8rcsjIyMcOHCBaSnp+Pp06coLCzEtGnT8OzZMwwfPhxxcXFIS0vD0aNHMXbsWJmTOzs7O9ja2sLNzQ3Hjx/HnTt3cOTIEWEF8Hnz5uHcuXOYPn06EhISkJKSgj/++KPK+4Ht7e0RGhoqJL6ampowMzNDeHj4eyfeM2bMQFBQEIKDg3H79m0sXrwY169flyozd+5cBAYGIjw8HLdu3cL8+fORkJCAGTNmSMV448YNXL9+HV9//bVU3F26dKlyBPpdJVO9vby8sH//fty5cwcXL15EQEAAIiIiABRfx9zcXERFReHp06d49eqVzO3Ler09PT0RERGBoKAgYVG1EiYmJti/fz8SEhKQmJiIESNG1OpMAiIiIiIqxsSbqJaoqKige/fuWLduHWxtbWFhYQE/Pz9MnDgRmzZtAlA8pfnPP/+Era0txo4dC1NTU3h4eODu3bvQ1dWV+Vg+Pj6Ql5eHubk5dHR0hAW8YmNjIZFI8O2336JDhw6YOXMmNDQ0qvU86X379qFr164YPnw4zM3N4evrKyRylpaWiImJwe3bt9GzZ09YWVlh0aJFVU6ptLOzg0QikbpP2t7evsy2mhg2bBj8/Pzg6+uLzp074+7du5g6dapUGW9vb8yePRtz5sxBhw4dEBkZiUOHDklNce/QoQM0NDTQqVMnYSr4+8YYHBwMLy8vzJkzB23btsXAgQMRFxcnTMu3sbHBlClTMGzYMOjo6GDVqlUyty3r9XZwcICmpiZu3bqFESNGSLWxdu1aNG3aFDY2NnB1dYWTk5PUiDgRERER1Q5RUVFRUX0HQUREn4bs7Gyoq6vj5cuXUFNTq+9wqA6kp6fD398f/v7+H3Q6dclxHY0c0VS5aY3aeP76OY6nH3+vNmRpv676pr76noiIyled7z0c8SYiIiIiIiKqQ0y8iYiIiIiIiOoQE28iIiIiIiKiOsTEm4iIiGSmr68Pf39/4Xnx9OGw74mIGi4+x5uIiIhkpqioyIW96gn7noio4eKINxEREREREVEdYuJNREREREREVIeYeBMRERERERHVISbeRERERERERHWIiTcRERERERFRHeKq5kRERNRgZOdn17xuXrbUf2vb+8RGRESfNibeRERE9NFTVVWFuJEYFx5ceO+2LmS+fxsVETcSQ1VVtc7aJyKihklUVFRUVN9BEBHRpyE7Oxvq6up4+fIl1NTU6jsc+sRkZWUhJyenvsOolKqqKrS0tOo7DCIi+gCq872HI95ERETUIGhpaTGpJSKiBomLqxERERERERHVISbeRERERERERHWIiTcRERERERFRHWLiTURERERERFSHmHgTERERERER1SGuak5EREQAGsbjut7Fx3cREVFDwMSbiIiIkJWVhQULFiA/P7++Q6kWsViMgIAAJt9ERPRRY+JNREREyMnJQX5+PpyH9oZmM80Pcsxnj58jcu9xOA91hGazpjWo/wyRe08gJyeHiTcREX3UmHgTERGRQLOZJpo11/nAx2z6wY9JRET0IXFxNSIiIiIiIqI6xMSbiIiIiIiIqA4x8SYiIiIiIiKqQ0y8iYiIiIiIiOoQE28iIiIiIiKiOsTEm4iIqJbk5eUhPT0deXl59R0KfYL4/iIiariYeBMREdWSzMxM+Pv7IzMzs75DoU8Q319ERA0XE28iIiIiIiKiOsTEm4iIiIiIiKgOMfGmBic9PR0ikQgJCQnv1Y6/vz86depUKzF9LGqrb6rLyMgI69ev/6DHrImQkBBoaGjUdxgftTFjxmDgwIH1HQYRERHRJ4WJdz148uQJpk6dipYtW0JRURF6enpwcnJCbGxsrR7H3t4eM2fOrNU2PwaGhobIzMyEhYVFfYfyQVQnWfyU+yY7OxsLFy5Eu3btoKSkBD09PfTu3Rv79+9HUVFRfYdHRERERFQhhfoO4HPk5uaG/Px87Nq1C61bt8ajR48QFRWFrKys+g6tQZCXl4eenl59h/HeJBIJRCIR5ORq5/ev/Px8iMXiBtE3Y8aMgZGREfz9/WUq/+LFC3z99dd4+fIlli1bhq5du0JBQQExMTHw9fWFg4NDvY5kFxQUoFGjRu/dTsk1/FD1iIiIiOjD4Ij3B/bixQucOXMGgYGB6NWrF1q1aoVu3bphwYIF6N+/v1S5CRMmQEdHB2pqanBwcEBiYqKwv2Sa9O7du2FkZAR1dXV4eHggJycHQHFiExMTgw0bNkAkEkEkEiE9PR0AcO3aNbi4uEBFRQW6uroYNWoUnj59KrRtb28Pb29v+Pr6QlNTE3p6emUSpBcvXmDy5MnQ1dWFkpISLCwscPjwYWH/2bNn0bNnTygrK8PQ0BDe3t74559/yu2Tly9fQl5eHvHx8QCAwsJCaGpq4ssvvxTK/PLLLzA0NARQdjp1dHQ0RCIRoqKi0KVLFzRu3Bg2Nja4deuW1HFWrlwJXV1dqKqqYvz48Xjz5o3U/sLCQixduhQtWrSAoqIiOnXqhMjISGH/kCFDMH36dOH1zJkzIRKJcPPmTQDFyU+TJk1w4sSJcs+zZOT60KFDMDc3h6KiIjIyMpCXlwcfHx80b94cTZo0Qffu3REdHS2c29ixY/Hy5UvhOpZcCyMjI/z444/w8vKCmpoaJk2aVO5U88qu9/bt22FgYIDCwkKpWAcMGIBx48YBANLS0jBgwADo6upCRUUFXbt2rfAc68q//vUvpKen48KFCxg9ejTMzc1hamqKiRMnIiEhASoqKgCA58+fw8vLC02bNkXjxo3h4uKClJSUMu0dPXoUZmZmUFFRgbOzc5kVgnfu3AkzMzMoKSmhXbt22LJli7CvpI/Dw8NhZ2cHJSUlhIaGVlmvPPb29pg+fTpmzpwJbW1tODk5AZDtM1pevbVr16JDhw5o0qQJDA0N8d133yE3N1eoV/IerOr8S4uLi4OOjg4CAwMrPRciIiIiqhgT7w9MRUUFKioqOHjwYKXP4XR3d8fjx49x5MgRXLp0CdbW1vjmm2/w7NkzoUxaWhoOHjyIw4cP4/Dhw4iJicHKlSsBABs2bECPHj0wceJEZGZmIjMzE4aGhnjx4gUcHBxgZWWF+Ph4REZG4tGjRxg6dKjU8Xft2oUmTZrgwoULWLVqFZYuXYrjx48DKE5QXVxcEBsbi19++QU3btzAypUrIS8vL8Tl7OwMNzc3JCUlITw8HGfPnpVKWktTV1dHp06dhGTz6tWrEIlEuHLlipA0xMTEwM7OrtK+XbhwIdasWYP4+HgoKCgIiSMA7N27F/7+/lixYgXi4+Ohr69fJinasGED1qxZg9WrVyMpKQlOTk7o37+/kLjZ2dkJMZbEpK2tLWyLi4tDQUEBbGxsKozx1atXCAwMxM6dO3H9+nU0a9YM06dPx19//YWwsDAkJSXB3d0dzs7OSElJgY2NDdavXw81NTXhOvr4+AjtrV69Gh07dsSVK1fg5+dX5nhVXW93d3dkZWXh1KlTQp1nz54hMjISnp6eAIDc3Fz06dMHUVFRuHLlCpydneHq6oqMjIxKr0dtKSwsRFhYGDw9PWFgYFBmv4qKChQUiifvjBkzBvHx8Th06BD++usvFBUVoU+fPigoKBDKv3r1CqtXr8bu3btx+vRpZGRkSPVpaGgoFi1ahOXLlyM5ORkrVqyAn58fdu3aJXXc+fPnY8aMGUhOToaTk5PM9d61a9cuiMVixMbGYtu2bdX6jJauBwBycnLYuHEjrl+/jl27duHkyZPw9fWVqlfV+Zd28uRJODo6Yvny5Zg3b16l50FEREREFeNU8w9MQUEBISEhmDhxIrZt2wZra2vY2dnBw8MDlpaWAIpHiy9evIjHjx9DUVERQHGCdfDgQfz++++YNGkSgOKEJCQkBKqqqgCAUaNGISoqCsuXL4e6ujrEYjEaN24sNfV406ZNsLKywooVK4RtQUFBMDQ0xO3bt2FqagoAsLS0xOLFiwEAJiYm2LRpE6KiouDo6IgTJ07g4sWLSE5OFsq3bt1aaC8gIACenp7C/eUmJibYuHEj7OzssHXrVigpKZXpF3t7e0RHR8PHxwfR0dFwdHTEzZs3cfbsWTg7OyM6OrpMAvGu5cuXC8n5/Pnz0bdvX7x58wZKSkpYv349xo8fj/HjxwMAli1bhhMnTkiNeq9evRrz5s2Dh4cHACAwMBCnTp3C+vXrsXnzZtjb22PGjBl48uQJFBQUcOPGDfj5+SE6OhpTpkxBdHQ0unbtisaNG1cYY0FBAbZs2YKOHTsCADIyMhAcHIyMjAwhqfTx8UFkZCSCg4OxYsUKqKurQyQSlTuF3MHBAXPmzBFel8xqKCHL9XZxccGePXvwzTffAAB+//13aGtro1evXgCAjh07CvECwI8//ogDBw7g0KFDFf6YUpuePn2K58+fo127dpWWS0lJwaFDhxAbGyv8+BEaGgpDQ0McPHgQ7u7uAIqvwbZt29CmTRsAwPTp07F06VKhncWLF2PNmjUYPHgwAOCLL77AjRs38J///AejR48Wys2cOVMoU5167zIxMcGqVauE18uWLZPpM/puvZKYShgZGWHZsmWYMmWK1I9MVZ1/iQMHDsDLyws7d+7EsGHDKow/Ly9P6kfE7OzsCst+Th48eFDfIVRbQ4y5REOOvTo+l/MkIvoUMfGuB25ubujbty/OnDmD8+fP48iRI1i1ahV27tyJMWPGIDExEbm5udDS0pKq9/r1a6SlpQmvjYyMhKQbAPT19fH48eNKj52YmIhTp04JU3NLS0tLk0q8SyvddkJCAlq0aCGULe8YSUlJwvRbACgqKkJhYSHu3LkDMzOzMnXs7Ozw888/QyKRICYmBt9++y309PQQHR0NS0tLpKamwt7evtJzKx2zvr4+AODx48do2bIlkpOTMWXKFKnyPXr0EEZ6s7Oz8eDBA3z11VdSZb766ithir+FhQU0NTURExMDsVgMKysr9OvXD5s3bwZQPAJeVYxisVgqzqtXr0IikZTpy7y8vDLXvzxdunSpdL8s19vT0xMTJ07Eli1boKioiNDQUHh4eAj3nufm5sLf3x8RERHIzMzE27dv8fr162qNeIeGhmLy5MlS5ycSibB69Wph25EjR9CzZ88ydWVdOC05ORkKCgro3r27sE1LSwtt27ZFcnKysK1x48ZC0glIv7f/+ecfpKWlYfz48Zg4caJQ5u3bt1BXV5c6Xum+r069d3Xu3Fnqtayf0XfrAcCJEycQEBCAmzdvIjs7G2/fvsWbN2/w6tUr4Qehys6/xIULF3D48GH8/vvvVa5wHhAQgCVLllRa5nO0ffv2+g7hs8L+JiKijx0T73qipKQER0dHODo6ws/PDxMmTMDixYsxZswY5ObmQl9fX2pac4nSC0i9u5iTSCQqc6/uu3Jzc+Hq6lru/ZolyWpVbSsrK1d5jMmTJ8Pb27vMvpYtW5Zbx9bWFjk5Obh8+TJOnz6NFStWQE9PDytXrkTHjh1hYGAAExOTSo9bOmaRSAQAVfZHdYhEItja2iI6OhqKioqwt7eHpaUl8vLycO3aNZw7d67CKbsllJWVhdiA4r6Sl5fHpUuXhKn6JcpLvN7VpEmTSvfLcr1dXV1RVFSEiIgIdO3aFWfOnMG6deuEcj4+Pjh+/DhWr14NY2NjKCsrY8iQIcjPz68yvhL9+/eXSojnzZuH5s2bS71HmjdvXm5dHR0daGhoCPfSv6/y3tslyX3JrQ07duyQihdAmetTuu+rU+9d715DWT+j79ZLT09Hv379MHXqVCxfvhyampo4e/Ysxo8fj/z8fCHxruz8S7Rp0wZaWloICgpC3759K104bsGCBZg9e7bwOjs7W1iP4XM2adKkcm+N+Jg9ePCgwSawDbG/a6IhXyMios8dE++PhLm5OQ4ePAgAsLa2xsOHD6GgoAAjI6MatykWiyGRSKS2WVtbY9++fTAyMhLui60uS0tL3Lt3T2ra67vHuHHjBoyNjWVuU0NDA5aWlti0aRMaNWqEdu3aoVmzZhg2bBgOHz5c5f3dVTEzM8OFCxfg5eUlbDt//rzwdzU1NRgYGCA2NlbqWLGxsejWrZvw2s7ODjt27ICioiKWL18OOTk52Nra4qeffkJeXl6ZEfOqWFlZQSKR4PHjx+WO9gLlX0dZyXK9lZSUMHjwYISGhiI1NRVt27aFtbW1sD82NhZjxozBoEGDABQnhu9Oaa+Kqqqq1OwMVVVVaGpqyvQekZOTg4eHB3bv3o3FixeX+XKdm5sLJSUlmJmZ4e3bt7hw4YIw1TwrKwu3bt2Cubm5THHq6urCwMAAf//9t3CPe13WK09NP6OXLl1CYWEh1qxZI8xW2Lt3b41i0NbWxv79+2Fvb4+hQ4di7969FSbfioqKwi0x9P8YGBi817/fVD3sbyIi+thxcbUPLCsrCw4ODvjll1+QlJSEO3fu4LfffsOqVaswYMAAAEDv3r3Ro0cPDBw4EMeOHUN6ejrOnTuHhQsXCit/y8LIyAgXLlxAeno6nj59isLCQkybNg3Pnj3D8OHDERcXh7S0NBw9ehRjx46VObmzs7ODra0t3NzccPz4cdy5cwdHjhwRVgCfN28ezp07h+nTpyMhIQEpKSn4448/qrwf2N7eHqGhoULiq6mpCTMzM2H16PcxY8YMBAUFITg4GLdv38bixYtx/fp1qTJz585FYGAgwsPDcevWLcyfPx8JCQmYMWOGVIw3btzA9evX8fXXX0vF3aVLlypHoN9VMtXby8sL+/fvx507d3Dx4kUEBAQgIiICQPF1zM3NRVRUFJ4+fYpXr17J3L6s19vT0xMREREICgoqkziamJhg//79SEhIQGJiIkaMGFGrMwlksXz5chgaGqJ79+7473//ixs3biAlJQVBQUGwsrJCbm4uTExMMGDAAEycOBFnz55FYmIiRo4ciebNmwufLVksWbIEAQEB2LhxI27fvo2rV68iODgYa9eurZN676rpZ9TY2BgFBQX497//jb///hu7d+8WFl2riWbNmuHkyZO4efMmhg8fjrdv39a4LSIiIqLPHRPvD0xFRQXdu3fHunXrYGtrCwsLC/j5+WHixInYtGkTgOKpn3/++SdsbW0xduxYmJqawsPDA3fv3oWurq7Mx/Lx8YG8vDzMzc2ho6MjLOAVGxsLiUSCb7/9Fh06dMDMmTOhoaFRredJ79u3D127dsXw4cNhbm4OX19fISmwtLRETEwMbt++jZ49e8LKygqLFi2qchqgnZ0dJBKJ1H3S9vb2ZbbVxLBhw+Dn5wdfX1907twZd+/exdSpU6XKeHt7Y/bs2ZgzZw46dOiAyMhIHDp0SGqKe4cOHaChoYFOnToJU8HfN8bg4GB4eXlhzpw5aNu2LQYOHIi4uDhhWr6NjQ2mTJmCYcOGQUdHp8yCWpWR9Xo7ODhAU1MTt27dwogRI6TaWLt2LZo2bQobGxu4urrCyclJakT8Q9DU1MT58+cxcuRIYfGxnj174tdff8VPP/0k3EcdHByMzp07o1+/fujRoweKiorw559/VusZ2xMmTMDOnTsRHByMDh06wM7ODiEhIfjiiy/qpN67avoZ7dixI9auXYvAwEBYWFggNDQUAQEB1Tr2u/T09HDy5ElcvXoVnp6eNZ55QURERPS5ExXJunIRERFRFbKzs6Guro6XL19CTU2tvsP54NLT0+Hv7w9/f/8GN/W5JPYR04eiWXOdD3LMx/efYM+mvTU+Zkn9htjfNdGQ319ERJ+i6nzv4Yg3ERERERERUR1i4k1ERERERERUh5h4ExEREREREdUhJt5ERES1RF9fH/7+/lLPXCeqLXx/ERE1XHyONxERUS1RVFTkoldUZ/j+IiJquDjiTURERERERFSHmHgTERERERER1SEm3kRERERERER1iIk3ERERERERUR1i4k1ERERERERUh7iqOREREQmePX72AY/1XOq/1a//4WIlIiJ6H0y8iYiICKqqqhCLxYjce+KDHzty7/Ea1xWLxVBVVa3FaIiIiGqfqKioqKi+gyAiok9DdnY21NXV8fLlS6ipqdV3OFRNWVlZyMnJqe8wqkVVVRVaWlr1HQYREX2GqvO9hyPeREREBADQ0tJiEktERFQHuLgaERERERERUR1i4k1ERERERERUh5h4ExEREREREdUhJt5EREREREREdYiJNxEREREREVEd4qrmRERE9aQhPr5LVnzMFxER0f/DxJuIiKgeZGVlYcGC+cjPL6jvUOqEWNwIAQErmXwTERGBiTcREVG9yMnJQX5+ATzddKCrLa6y/KMn+Qjd/wSeg3Wgq1N1+fr06Gk+Qvc9QU5ODhNvIiIiMPEmIiKqV7raYrQwUJS9vE71yhMREVH94+JqRERERERERHWIiTcRERERERFRHWLiTURERERERFSHmHgTERERERER1SEm3kRERERERER1iIk3ERE1OHl5eUhPT0deXl59h0J1jNeaiIg+BUy8iYiowcnMzIS/vz8yMzPrOxSqY7zWRET0KWDiTURERERERFSHmHgTERERERER1SEm3vTZSk9Ph0gkQkJCwnu14+/vj06dOtVKTB+L2uqb6jIyMsL69es/6DHp/6mv605ERET0qWPi3YA9efIEU6dORcuWLaGoqAg9PT04OTkhNja2Vo9jb2+PmTNn1mqbHwNDQ0NkZmbCwsKivkP5IEJCQqChoSFT2U+1b6r6kaSq97pIJIJIJML58+eltufl5UFLSwsikQjR0dEV1h8zZozQRqNGjfDFF1/A19cXb968qeaZENHnprCwEMnJyTh//jySk5NRWFhY3yEREVE1KNR3AFRzbm5uyM/Px65du9C6dWs8evQIUVFRyMrKqu/QGgR5eXno6enVdxjvTSKRQCQSQU6udn5Hy8/Ph1gsbhB9M2bMGBgZGcHf3/+DHdPQ0BDBwcH48ssvhW0HDhyAiooKnj17VmV9Z2dnBAcHo6CgAJcuXcLo0aMhEokQGBhYo3iKioogkUigoMB/zok+VfHx8QgLC8PTp0+Fbdra2vDw8ECXLl3qMTIiIpIVR7wbqBcvXuDMmTMIDAxEr1690KpVK3Tr1g0LFixA//79pcpNmDABOjo6UFNTg4ODAxITE4X9JSOAu3fvhpGREdTV1eHh4YGcnBwAxYlNTEwMNmzYIIzUpaenAwCuXbsGFxcXqKioQFdXF6NGjZL6UmBvbw9vb2/4+vpCU1MTenp6ZRKkFy9eYPLkydDV1YWSkhIsLCxw+PBhYf/Zs2fRs2dPKCsrw9DQEN7e3vjnn3/K7ZOXL19CXl4e8fHxAIpHBzQ1NaUSpF9++QWGhoYAyk6rjY6OhkgkQlRUFLp06YLGjRvDxsYGt27dkjrOypUroaurC1VVVYwfP77MaGVhYSGWLl2KFi1aQFFREZ06dUJkZKSwf8iQIZg+fbrweubMmRCJRLh58yaA4sS3SZMmOHHiRLnnWTJyfejQIZibm0NRUREZGRnIy8uDj48PmjdvjiZNmqB79+7C6Gt0dDTGjh2Lly9fCtex5FoYGRnhxx9/hJeXF9TU1DBp0qRypxxXdr23b98OAwODMiMwAwYMwLhx4wAAaWlpGDBgAHR1daGiooKuXbtWeI4fs9GjRyMsLAyvX78WtgUFBWH06NEy1S+ZnWJoaIiBAweid+/eOH78uLC/sLAQAQEB+OKLL6CsrIyOHTvi999/F/aXvE+PHDmCzp07Q1FREWfPnpWpf42MjLBixQqMGzcOqqqqaNmyJbZv315hrBKJBOPGjUO7du2QkZEhaxcRUS2Kj4/H5s2b0aJFC/zwww/Ytm0bfvjhB7Ro0QKbN28W/p9HREQfNybeDZSKigpUVFRw8ODBSp9t6u7ujsePH+PIkSO4dOkSrK2t8c0330iNzKWlpeHgwYM4fPgwDh8+jJiYGKxcuRIAsGHDBvTo0QMTJ05EZmYmMjMzYWhoiBcvXsDBwQFWVlaIj49HZGQkHj16hKFDh0odf9euXWjSpAkuXLiAVatWYenSpUKSUVhYCBcXF8TGxuKXX37BjRs3sHLlSsjLywtxOTs7w83NDUlJSQgPD8fZs2elktbS1NXV0alTJyHZvHr1KkQiEa5cuYLc3FwAQExMDOzs7Crt24ULF2LNmjWIj4+HgoKCkDgCwN69e+Hv748VK1YgPj4e+vr62LJli1T9DRs2YM2aNVi9ejWSkpLg5OSE/v37IyUlBQBgZ2cnNR05JiYG2trawra4uDgUFBTAxsamwhhfvXqFwMBA7Ny5E9evX0ezZs0wffp0/PXXXwgLC0NSUhLc3d3h7OyMlJQU2NjYYP369VBTUxOuo4+Pj9De6tWr0bFjR1y5cgV+fn5ljlfV9XZ3d0dWVhZOnTol1Hn27BkiIyPh6ekJAMjNzUWfPn0QFRWFK1euwNnZGa6urg0uoevcuTOMjIywb98+AEBGRgZOnz6NUaNGVbuta9eu4dy5cxCLxcK2gIAA/Pe//8W2bdtw/fp1zJo1CyNHjkRMTIxU3fnz52PlypVITk6GpaWlzP27Zs0adOnSBVeuXMF3332HqVOnlvlxCSiePu/u7o6EhAScOXMGLVu2rPb5EdH7KSwsRFhYGDp27Ahvb28YGxtDSUkJxsbG8Pb2RseOHREeHs5p50REDQDnJjZQCgoKCAkJwcSJE7Ft2zZYW1vDzs4OHh4esLS0BFA8Wnzx4kU8fvwYioqKAIoTrIMHD+L333/HpEmTABT/jz0kJASqqqoAgFGjRiEqKgrLly+Huro6xGIxGjduLDX1eNOmTbCyssKKFSuEbUFBQTA0NMTt27dhamoKALC0tMTixYsBACYmJti0aROioqLg6OiIEydO4OLFi0hOThbKt27dWmgvICAAnp6ewj23JiYm2LhxI+zs7LB161YoKSmV6Rd7e3tER0fDx8cH0dHRcHR0xM2bN3H27Fk4OzsjOjoavr6+lfbt8uXLheR8/vz56Nu3L968eQMlJSWsX78e48ePx/jx4wEAy5Ytw4kTJ6RGvVevXo158+bBw8MDABAYGIhTp05h/fr12Lx5M+zt7TFjxgw8efIECgoKuHHjBvz8/BAdHY0pU6YgOjoaXbt2RePGjSuMsaCgAFu2bEHHjh0BFCd/wcHByMjIgIGBAQDAx8cHkZGRCA4OxooVK6Curg6RSFTuFHIHBwfMmTNHeF0yq6GELNfbxcUFe/bswTfffAMA+P3336GtrY1evXoBADp27CjECwA//vgjDhw4gEOHDlX4Y8rHaty4cQgKCsLIkSMREhKCPn36QEdHR6a6hw8fhoqKCt6+fYu8vDzIyclh06ZNAIqT3RUrVuDEiRPo0aMHgOLPxNmzZ/Gf//xH6kejpUuXwtHRUXitqakpU//26dMH3333HQBg3rx5WLduHU6dOoW2bdsKZXJzc9G3b1/k5eXh1KlTUFdXr/B88vLypH78y87OlqkfasuDBw8+6PFqU0OOXVa1cY6fQz9V5NatW3j69CmmTJlS5nYiOTk59OvXD8uWLcOtW7dgZmZWT1ESEZEsmHg3YG5ubujbty/OnDmD8+fP48iRI1i1ahV27tyJMWPGIDExEbm5udDS0pKq9/r1a6SlpQmvjYyMhKQbAPT19fH48eNKj52YmIhTp05BRUWlzL60tDSpxLu00m0nJCSgRYsWQtnyjpGUlITQ0FBhW1FREQoLC3Hnzp1yv2TY2dnh559/hkQiQUxMDL799lvo6ekhOjoalpaWSE1Nhb29faXnVjpmfX19AMDjx4/RsmVLJCcnY8qUKVLle/ToIYz0Zmdn48GDB/jqq6+kynz11VfCFH8LCwtoamoiJiYGYrEYVlZW6NevHzZv3gygeAS8qhjFYrFUnFevXoVEIinTlyWLflWlqnsEZbnenp6emDhxIrZs2QJFRUWEhobCw8ND+LKYm5sLf39/REREIDMzE2/fvsXr16+rNeIdGhqKyZMnS52fSCTC6tWrhW1HjhxBz549ZW6zJkaOHIn58+fj77//RkhICDZu3Chz3V69emHr1q34559/sG7dOigoKMDNzQ0AkJqailevXkkl1EDx7QdWVlZS2969ZrL2b+n3TckPMe9+3ocPH44WLVrg5MmTUFZWrvR8AgICsGTJEtlOvg5UNlWe6h+vz/t5+fIlAKBFixbl7m/evLlUOSIi+ngx8W7glJSU4OjoCEdHR/j5+WHChAlYvHgxxowZg9zcXOjr65e7ynLp1a0bNWoktU8kElU5bS03Nxeurq7lLghVkqxW1XZVX+hzc3MxefJkeHt7l9lX0bRXW1tb5OTk4PLlyzh9+jRWrFgBPT09rFy5Eh07doSBgQFMTEwqPW7pmEUiEQDU6jQ+kUgEW1tbREdHQ1FREfb29rC0tEReXp4w9bj0NPDyKCsrC7EBxX0lLy+PS5cuCVP1S5SXLL+rSZMmle6X5Xq7urqiqKgIERER6Nq1K86cOYN169YJ5Xx8fHD8+HGsXr0axsbGUFZWxpAhQ5Cfn19lfCX69++P7t27C6/nzZuH5s2bS71HSr6I1iUtLS3069dPuMffxcVFWBehKk2aNIGxsTGA4lkDHTt2xM8//4zx48cLt0RERESUOY+SWSul2ylN1v6V5fPep08f/PLLL/jrr7/g4OBQ6fksWLAAs2fPFl5nZ2cL6yh8CJMmTRJmeTQ0Dx48+OQT09q4Pp9DP1WkZLbJvXv3hH83Srt//75UOSIi+ngx8f7EmJub4+DBgwAAa2trPHz4EAoKCjAyMqpxm2KxGBKJRGqbtbU19u3bByMjoxqvpmxpaYl79+5JTU1/9xg3btwo98tGRTQ0NGBpaYlNmzahUaNGaNeuHZo1a4Zhw4bh8OHDVd7fXRUzMzNcuHABXl5ewrbSj5ZSU1ODgYEBYmNjpY4VGxuLbt26Ca/t7OywY8cOKCoqYvny5ZCTk4OtrS1++ukn5OXllRkxr4qVlRUkEgkeP35c4WhveddRVrJcbyUlJQwePBihoaFITU1F27ZtYW1tLeyPjY3FmDFjMGjQIADFyfy7U9qroqqqKjU7Q1VVFZqamtV6j9SWcePGoU+fPpg3b16ZHztkJScnh3/961+YPXs2RowYIbVYXnXfq7XRvyWmTp0KCwsL9O/fHxEREZXGoqioWOZHgQ/JwMDgvf59o7rF6/N+2rZtC21tbRw+fBje3t5S080LCwtx+PBh6OjoSN0qQkREHycurtZAZWVlwcHBAb/88guSkpJw584d/Pbbb1i1ahUGDBgAAOjduzd69OiBgQMH4tixY0hPT8e5c+ewcOHCaq2CamRkhAsXLiA9PR1Pnz5FYWEhpk2bhmfPnmH48OGIi4tDWloajh49irFjx8qc3NnZ2cHW1hZubm44fvw47ty5gyNHjggrgM+bNw/nzp3D9OnTkZCQgJSUFPzxxx9V3g9sb2+P0NBQIVnQ1NSEmZkZwsPD3zvxnjFjBoKCghAcHIzbt29j8eLFuH79ulSZuXPnIjAwEOHh4bh16xbmz5+PhIQEzJgxQyrGGzdu4Pr16/j666+l4u7SpUuVI9DvKpnq7eXlhf379+POnTu4ePEiAgICEBERAaD4Oubm5iIqKgpPnz7Fq1evZG5f1uvt6emJiIgIBAUFCYuqlTAxMcH+/fuRkJCAxMREjBgxol4WBHr9+jUSEhKk/pS+9eLJkydl9j969KhMO87Oznjy5AmWLl36XvG4u7tDXl4emzdvhqqqKnx8fDBr1izs2rULaWlpuHz5Mv79739j165dlbZT2/37/fffY9myZejXrx/Onj1b43aIqObk5OTg4eGBxMREbNy4EampqXj9+jVSU1OxceNGJCYmYtiwYbX2OEkiIqo7/Je6gVJRUUH37t2xbt062NrawsLCAn5+fpg4caKwUJNIJMKff/4JW1tbjB07FqampvDw8MDdu3ehq6sr87F8fHwgLy8Pc3Nz6OjoCAt4xcbGQiKR4Ntvv0WHDh0wc+ZMaGhoVOsLwL59+9C1a1cMHz4c5ubm8PX1FRI5S0tLxMTE4Pbt2+jZsyesrKywaNGiKqct2tnZQSKRSN0nbW9vX2ZbTQwbNgx+fn7w9fVF586dcffuXUydOlWqjLe3N2bPno05c+agQ4cOiIyMxKFDh6SmuHfo0AEaGhro1KmTMBX8fWMMDg6Gl5cX5syZg7Zt22LgwIGIi4sTpuXb2NhgypQpGDZsGHR0dLBq1SqZ25b1ejs4OEBTUxO3bt3CiBEjpNpYu3YtmjZtChsbG7i6usLJyUlqRPxDuX37NqysrKT+lL5vfM+ePWX279ixo0w7IpEI2traUiuS14SCggKmT5+OVatW4Z9//sGPP/4IPz8/BAQEwMzMDM7OzoiIiMAXX3xRaTt10b8zZ87EkiVL0KdPH5w7d+692iKimunSpQumTZuGe/fuYdmyZZg6dSqWLVuG+/fvY9q0aXyONxFRAyEqKioqqu8giIjo05CdnQ11dXW8fPkSampqdXac9PR0+Pv7w9/fv8FOZS45h9mTm6OFQdXT9e89yMPa/9yXuXx9Kom1Nq7Pp3Cta0NhYSFu3bqFly9fQl1dHW3btuVINxFRPavO9x7e401ERET0kZOTk+Mjw4iIGjD+VEpERERERERUh5h4ExEREREREdUhJt5ERNTg6Ovrw9/fX3iOPH26eK2JiOhTwHu8iYiowVFUVPysF9r6nPBaExHRp4Aj3kRERERERER1iIk3ERERERERUR1i4k1ERERERERUh5h4ExEREREREdUhJt5EREREREREdYirmhMREdWjR0/zZSv3JF/qvx8zWc+JiIjoc8HEm4iIqB6oqqpCLG6E0H1PqlUvdH/1ytcXsbgRVFVV6zsMIiKij4KoqKioqL6DICKiT0N2djbU1dXx8uVLqKmp1Xc4H72srCzk5OTUdxh1QlVVFVpaWvUdBhERUZ2pzvcejngTERHVEy0tLSanREREnwEurkZERERERERUh5h4ExEREREREdUhJt5EREREREREdYiJNxEREREREVEdYuJNREREREREVIe4qjkRERF9ND7lR6zJgo9hIyL6NDHxJiIioo9CVlYW5i9YgIL8/PoOpd40EouxMiCAyTcR0SeGiTcRERF9FHJyclCQn482ve2grKle6+2/fv4CacdPo42jLZSbatR6++/r9bOXSDsRg5ycHCbeRESfGCbeRERE9FFR1lRHEx3tumu/qUadtk9ERPQuLq5GREREREREVIeYeBMRERERERHVISbeRERERERERHWIiTcRERERERFRHWLiTURERERERFSHmHgTERGRzPLy8pCeno68vLz6DoUaEL5viOhzx8SbiIiIZJaZmQl/f39kZmbWdyjUgPB9Q0SfOybeRERERERERHWIiTcRERERERFRHWLiTdQApKenQyQSISEh4b3a8ff3R6dOnWolpo9FbfVNdRkZGWH9+vUf9JhERERE1DAx8aYP4smTJ5g6dSpatmwJRUVF6OnpwcnJCbGxsbV6HHt7e8ycObNW2/wYGBoaIjMzExYWFvUdygcREhICDQ0Nmcp+qn1T0Y8kNfmhwcjICCKRCCKRCMrKyjAyMsLQoUNx8uTJctsu78/58+ff84yIiIiIPl9MvOmDcHNzw5UrV7Br1y7cvn0bhw4dgr29PbKysuo7tAZBXl4eenp6UFBQqO9Q3otEIkFhYWGttZefn99g+mbMmDHw9/evt+MvXboUmZmZuHXrFv773/9CQ0MDvXv3xvLly8uUPXHiBDIzM6X+dO7cuR6iJiIiIvo0MPGmOvfixQucOXMGgYGB6NWrF1q1aoVu3bphwYIF6N+/v1S5CRMmQEdHB2pqanBwcEBiYqKwv2QEcPfu3TAyMoK6ujo8PDyQk5MDoDixiYmJwYYNG4RRuvT0dADAtWvX4OLiAhUVFejq6mLUqFF4+vSp0La9vT28vb3h6+sLTU1N6OnplUmSXrx4gcmTJ0NXVxdKSkqwsLDA4cOHhf1nz55Fz549oaysDENDQ3h7e+Off/4pt09evnwJeXl5xMfHAwAKCwuhqamJL7/8Uijzyy+/wNDQEEDZUc7o6GiIRCJERUWhS5cuaNy4MWxsbHDr1i2p46xcuRK6urpQVVXF+PHj8ebNG6n9hYWFWLp0KVq0aAFFRUV06tQJkZGRwv4hQ4Zg+vTpwuuZM2dCJBLh5s2bAIoT3yZNmuDEiRPlnmfJyPWhQ4dgbm4ORUVFZGRkIC8vDz4+PmjevDmaNGmC7t27Izo6Wji3sWPH4uXLl8J1LLkWRkZG+PHHH+Hl5QU1NTVMmjSp3BHgyq739u3bYWBgUOYHgAEDBmDcuHEAgLS0NAwYMAC6urpQUVFB165dKzzH+lRUVARjY2OsXr1aantCQgJEIhFSU1OFbaqqqtDT00PLli1ha2uL7du3w8/PD4sWLSrzvtHS0oKenp7Un0aNGn2QcyIiIiL6FDHxpjqnoqICFRUVHDx4sNLnd7q7u+Px48c4cuQILl26BGtra3zzzTd49uyZUCYtLQ0HDx7E4cOHcfjwYcTExGDlypUAgA0bNqBHjx6YOHGiMEpnaGiIFy9ewMHBAVZWVoiPj0dkZCQePXqEoUOHSh1/165daNKkCS5cuIBVq1Zh6dKlOH78OIDiBNXFxQWxsbH45ZdfcOPGDaxcuRLy8vJCXM7OznBzc0NSUhLCw8Nx9uxZqaS1NHV1dXTq1ElINq9evQqRSIQrV64gNzcXABATEwM7O7tK+3bhwoVYs2YN4uPjoaCgICSOALB37174+/tjxYoViI+Ph76+PrZs2SJVf8OGDVizZg1Wr16NpKQkODk5oX///khJSQEA2NnZCTGWxKStrS1si4uLQ0FBAWxsbCqM8dWrVwgMDMTOnTtx/fp1NGvWDNOnT8dff/2FsLAwJCUlwd3dHc7OzkhJSYGNjQ3Wr18PNTU14Tr6+PgI7a1evRodO3bElStX4OfnV+Z4VV1vd3d3ZGVl4dSpU0KdZ8+eITIyEp6engCA3Nxc9OnTB1FRUbhy5QqcnZ3h6uqKjIyMSq/HhyYSiTBu3DgEBwdLbQ8ODoatrS2MjY0rrT9jxgwUFRXhjz/+qMswiYiIiD57H/fcTPokKCgoICQkBBMnTsS2bdtgbW0NOzs7eHh4wNLSEkDxaPHFixfx+PFjKCoqAihOsA4ePIjff/8dkyZNAlCcAIeEhEBVVRUAMGrUKERFRWH58uVQV1eHWCxG48aNoaenJxx/06ZNsLKywooVK4RtQUFBMDQ0xO3bt2FqagoAsLS0xOLFiwEAJiYm2LRpE6KiouDo6IgTJ07g4sWLSE5OFsq3bt1aaC8gIACenp7C/eUmJibYuHEj7OzssHXrVigpKZXpF3t7e0RHR8PHxwfR0dFwdHTEzZs3cfbsWTg7OyM6Ohq+vr6V9u3y5cuF5Hz+/Pno27cv3rx5AyUlJaxfvx7jx4/H+PHjAQDLli3DiRMnpEa9V69ejXnz5sHDwwMAEBgYiFOnTmH9+vXYvHkz7O3tMWPGDDx58gQKCgq4ceMG/Pz8EB0djSlTpiA6Ohpdu3ZF48aNK4yxoKAAW7ZsQceOHQEAGRkZCA4ORkZGBgwMDAAAPj4+iIyMRHBwMFasWAF1dXWIRCKp61jCwcEBc+bMEV6XzGooIcv1dnFxwZ49e/DNN98AAH7//Xdoa2ujV69eAICOHTsK8QLAjz/+iAMHDuDQoUMV/phSF65evQoVFRWpbUVFRVKvx4wZg0WLFuHixYvo1q0bCgoKsGfPnjKj4OXR1NREs2bNyvShjY0N5OSkf5ct+UHoXXl5eVI/qGVnZ1d5XPo0PHjwoEG02RB9iv3wKZ4TEVF1MPGmD8LNzQ19+/bFmTNncP78eRw5cgSrVq3Czp07MWbMGCQmJiI3NxdaWlpS9V6/fo20tDThtZGRkZB0A4C+vj4eP35c6bETExNx6tSpMgkMUDxSXTrxLq102wkJCWjRooVQtrxjJCUlITQ0VNhWVFSEwsJC3LlzB2ZmZmXq2NnZ4eeff4ZEIkFMTAy+/fZb6OnpITo6GpaWlkhNTYW9vX2l51Y6Zn19fQDA48eP0bJlSyQnJ2PKlClS5Xv06CGM9GZnZ+PBgwf46quvpMp89dVXwhR/CwsLaGpqIiYmBmKxGFZWVujXrx82b94MoHgEvKoYxWKxVJxXr16FRCIp05d5eXllrn95unTpUul+Wa63p6cnS31cbAAAXdtJREFUJk6ciC1btkBRURGhoaHw8PAQks3c3Fz4+/sjIiICmZmZePv2LV6/fl2tEe/Q0FBMnjxZ6vxEIpFUQnzkyBH07Nmzwjbatm2LQ4cOSW27f/++VJ8bGBigb9++CAoKQrdu3fB///d/yMvLg7u7u0xxFhUVQSQSSW0LDw8v9z1bnoCAACxZskSmsvRp2b59e32H8Mli3xIRfXqYeNMHo6SkBEdHRzg6OsLPzw8TJkzA4sWLMWbMGOTm5kJfX19qWnOJ0qtbv3ufqUgkqnKxrtzcXLi6uiIwMLDMvpJktaq2lZWVqzzG5MmT4e3tXWZfy5Yty61ja2uLnJwcXL58GadPn8aKFSugp6eHlStXomPHjjAwMICJiUmlxy0dc0nyVJuLl4lEItja2iI6OhqKioqwt7eHpaUl8vLycO3aNZw7d05qGnh5lJWVpRK73NxcyMvL49KlS8JU/RLlJcvvatKkSaX7Zbnerq6uKCoqQkREBLp27YozZ85g3bp1QjkfHx8cP34cq1evhrGxMZSVlTFkyBDk5+dXGV+J/v37o3v37sLrefPmoXnz5lLvkebNm1fahlgsLjNdvLxF5CZMmIBRo0Zh3bp1CA4OxrBhwyqdhVAiKysLT548wRdffCG13dDQsMpp6iUWLFiA2bNnC6+zs7OFtQno0zZp0iRh1kptefDgAZNO1E3f1jdeWyL63DHxpnpjbm6OgwcPAgCsra3x8OFDKCgowMjIqMZtisViSCQSqW3W1tbYt28fjIyMarzytaWlJe7duyc1Nf3dY9y4cUPmZAUo/kHB0tISmzZtQqNGjdCuXTs0a9YMw4YNw+HDh6u8v7sqZmZmuHDhAry8vIRtpR8JpaamBgMDA8TGxkodKzY2Ft26dRNe29nZYceOHVBUVMTy5cshJycHW1tb/PTTT8jLyyszYl4VKysrSCQSPH78uMLR3vKuo6xkud5KSkoYPHgwQkNDkZqairZt28La2lrYHxsbizFjxmDQoEEAipP5d6djV0VVVVVqdoaqqio0NTWr9R6RVZ8+fdCkSRNs3boVkZGROH36tEz1NmzYADk5OQwcOLDGx1ZUVBRuD6HPi4GBwXv9e00VY98SEX16uLga1bmsrCw4ODjgl19+QVJSEu7cuYPffvsNq1atwoABAwAAvXv3Ro8ePTBw4EAcO3YM6enpOHfuHBYuXCis/C0LIyMjXLhwAenp6Xj69CkKCwsxbdo0PHv2DMOHD0dcXBzS0tJw9OhRjB07Vubkzs7ODra2tnBzc8Px48dx584dHDlyRFgBfN68eTh37hymT5+OhIQEpKSk4I8//qjyfmB7e3uEhoYKia+mpibMzMwQHh7+3on3jBkzEBQUhODgYNy+fRuLFy/G9evXpcrMnTsXgYGBCA8Px61btzB//nwkJCRgxowZUjHeuHED169fx9dffy0Vd5cuXaocgX5XyVRvLy8v7N+/H3fu3MHFixcREBCAiIgIAMXXMTc3F1FRUXj69ClevXolc/uyXm9PT09EREQgKChIWFSthImJCfbv34+EhAQkJiZixIgRtTqToLbJy8tjzJgxWLBgAUxMTNCjR48yZXJycvDw4UP873//w+nTpzFp0iQsW7YMy5cvL/NjQFZWFh4+fCj1590V8YmIiIhIdky8qc6pqKige/fuWLduHWxtbWFhYQE/Pz9MnDgRmzZtAlA8pfnPP/+Era0txo4dC1NTU3h4eODu3bvQ1dWV+Vg+Pj6Ql5eHubk5dHR0hAW8YmNjIZFI8O2336JDhw6YOXMmNDQ0yiwgVZl9+/aha9euGD58OMzNzeHr6yskcpaWloiJicHt27fRs2dPWFlZYdGiRVVOFbSzs4NEIpG6Z9fe3r7MtpoYNmwY/Pz84Ovri86dO+Pu3buYOnWqVBlvb2/Mnj0bc+bMQYcOHfD/tXffYVFc7d/AvwtKb1KkCIpKtYANfcRHQMWuUWJBJQqKBUuwYYsF7CX2x96wRIMmmmhQNIqCiA0s2LBgxIoNI0WUOu8fvsyPFcRFWRbx+7muvXRnzpy5p+yy95wzZw4fPowDBw5IdXGvX78+9PT00KBBA7Er+JfGGBwcjAEDBmD8+PGwtbVF9+7dERMTI3bLd3Z2hp+fHzw9PWFkZIRFixbJXLesx7t169bQ19fHrVu30K9fP6k6li5diipVqsDZ2Rldu3ZF+/btpVrEyyNfX19kZWVh4MCBRc6fMWMGTE1NYWVlhf79+yMlJQXh4eGYNGlSobLu7u4wNTWVeuX3TiEiIiKikpMIHw6RS0REX52oqCi0adMGDx8+LNHFqtKWmpoKXV1dpKSkQEdHR2FxkPwkJiYiKCgIQUFBpd4dOr/uer2/g6aRYanWDQBvXrzEtT0H5Fb/l8qPTx77VtHked4QESlKSX738B5vIqKvWGZmJl68eIGgoCD06tVLoUk3ERERERWNXc2JiL5iv/76K2rUqIHXr1+XqEs+EREREZUdJt5ERF8xHx8f5Obm4sKFC598PBkRERERKQYTbyIiIpKZqakpgoKCYGpqquhQ6CvC84aIvnW8x5uIiIhkpqqqysGxqMR43hDRt44t3kRERERERERyxMSbiIiIiIiISI6YeBMRERERERHJERNvIiIiIiIiIjli4k1EREREREQkRxzVnIiIiMqVt69S5FPvv6+l/i1v5LXdRESkeEy8iYiIqFzQ1tZGZRUV3D0WKdf13D16Uq71f4nKKirQ1tZWdBhERFTKJIIgCIoOgoiIKobU1FTo6uoiJSUFOjo6ig6HvkLJyclIS0tTdBgKo62tDQMDA0WHQUREMijJ7x62eBMREVG5YWBgwMSTiIgqHA6uRkRERERERCRHTLyJiIiIiIiI5IiJNxEREREREZEcMfEmIiIiIiIikiMm3kRERERERERyxFHNiYiI6KtQXh41xkd+ERFRSTHxJiIionIvOTkZU6ZMQVZWlqJDgYqKCubPn8/km4iIZMbEm4iIiMq9tLQ0ZGVlYeDAFjAx0S00/+nTFAQHR390fmnJX09aWhoTbyIikhkTbyIiIvpqmJjoonr1jye8n5pPRESkCBxcjYiIiIiIiEiOmHgTERERERERyRETbyIiIiIiIiI5YuJNREREREREJEdMvImIiIiIiIjkiIk3ERFRBZeZmYnExERkZmYqOpQKIysri/uUiIhkxsSbiIiogktKSkJQUBCSkpIUHUqF8fLlS+5TIiKSGRNvIiIiIiIiIjli4k1EREREREQkR0y8v2KJiYmQSCS4fPnyF9UTFBSEBg0alEpM5UVp7ZuSsrS0xPLly8t0nVQ8HpOS8fHxQffu3RUdBhEREVGFotDE+8WLFxg+fDiqV68OVVVVmJiYoH379oiOji7V9bi5uWHMmDGlWmd5YGFhgaSkJNSrV0/RoZSJrVu3Qk9PT6ayFXXffOwiyedcaLC0tIREIoFEIoG6ujosLS3Ru3dvHD9+vMi6i3qdPXv2C7eoeCdOnECnTp1gYGAADQ0N1KlTB+PHj8fjx4/lul55+XBf6uvrw9XVFVFRUYoOjYhISl5eHuLj43H27FnEx8cjLy+vXNRFRPS1qqTIlffo0QNZWVnYtm0batWqhWfPniE8PBzJycmKDOuroaysDBMTE0WH8cVyc3MhkUigpFQ614GysrKgoqLyVewbHx8fWFpaIigoSCHrnzVrFoYMGSKOzvvLL7/A3d0ds2fPxtSpU6XKHjt2DHXr1pWaZmBgINN6EhMTUbNmTQiCIHNs69evx4gRI+Dt7Y29e/fC0tISDx48wPbt27FkyRIsXbpU5rpKW/459rny9+XLly8xd+5cdOnSBbdv34axsbFC4iEiKig2NhYhISF4+fKlOM3Q0BB9+vRBkyZNFFYXEdHXTGEt3q9fv0ZUVBQWLlyIVq1aoUaNGmjatCmmTJmC7777Tqrc4MGDYWRkBB0dHbRu3RpxcXHi/PwWwB07dsDS0hK6urro06cP0tLSALxPbCIjI7FixQqxlSkxMREAcO3aNXTs2BFaWlowNjZG//79pf4wuLm5wd/fHxMnToS+vj5MTEwKJUivX7/GsGHDYGxsDDU1NdSrVw+hoaHi/FOnTqFly5ZQV1eHhYUF/P398ebNmyL3SUpKCpSVlREbGwvg/RVifX19/Oc//xHL/PLLL7CwsABQuJUzIiICEokE4eHhaNKkCTQ0NODs7Ixbt25JrWfBggUwNjaGtrY2fH198e7dO6n5eXl5mDVrFszNzaGqqooGDRrg8OHD4vyePXti1KhR4vsxY8ZAIpHg5s2bAN4nAZqamjh27FiR25nfcn3gwAHUqVMHqqqqePDgATIzMxEQEIBq1apBU1MTzZo1Q0REhLhtAwcOREpKingc84+FpaUlZs+ejQEDBkBHRwdDhw4tsgW4uOO9YcMGmJmZFboK361bNwwaNAgAcPfuXXTr1g3GxsbQ0tKCk5PTR7dRkQRBgJWVFRYvXiw1/fLly5BIJEhISBCnaWtrw8TEBNWrV4eLiws2bNiA6dOnY8aMGYXOGwMDA5iYmEi9KleuLJdtePToEfz9/eHv748tW7bAzc0NlpaWcHFxwaZNmzBjxgyx7N69e1G3bl2oqqrC0tISS5YsKVRfRkYGBg0aBG1tbVSvXh0bNmyQmv/w4UP07t0benp60NfXR7du3cTvCeD/ul/PnTsXZmZmsLW1lWm5j8nfl/Xq1cNPP/2E1NRUnDt3Tpwvy3fTqFGjMGbMGBgaGqJ9+/YAgKVLl6J+/frQ1NSEhYUFRowYgfT0dHG5/M/ekSNHYG9vDy0tLXTo0KHYUZljYmJgZGSEhQsXfnK7iOjrFxsbi9WrV8Pc3BzTpk3DunXrMG3aNJibm2P16tXib5SyrouI6GunsMRbS0sLWlpa+PPPP4t9BmavXr3w/PlzhIWF4cKFC2jUqBHatGmDV69eiWXu3r2LP//8E6GhoQgNDUVkZCQWLFgAAFixYgWaN2+OIUOGICkpCUlJSbCwsMDr16/RunVrNGzYELGxsTh8+DCePXuG3r17S61/27Zt0NTUxLlz57Bo0SLMmjULR48eBfA+Qe3YsSOio6Pxyy+/4MaNG1iwYAGUlZXFuDp06IAePXrgypUr2L17N06dOiWVtBakq6uLBg0aiMnm1atXIZFIcOnSJfHHc2RkJFxdXYvdt1OnTsWSJUsQGxuLSpUqiYkjAOzZswdBQUGYN28eYmNjYWpqijVr1kgtv2LFCixZsgSLFy/GlStX0L59e3z33Xe4c+cOAMDV1VWMMT8mQ0NDcVpMTAyys7Ph7Oz80RgzMjKwcOFCbNq0CdevX0fVqlUxatQonDlzBiEhIbhy5Qp69eqFDh064M6dO3B2dsby5cuho6MjHseAgACxvsWLF8PR0RGXLl3C9OnTC63vU8e7V69eSE5OxokTJ8RlXr16hcOHD8PLywsAkJ6ejk6dOiE8PByXLl1Chw4d0LVrVzx48KDY41HWJBIJBg0ahODgYKnpwcHBcHFxgZWVVbHLjx49GoIgYP/+/fIMs1i//fYbsrKyMHHixCLn599ycOHCBfTu3Rt9+vTB1atXERQUhOnTp2Pr1q1S5ZcsWYImTZrg0qVLGDFiBIYPHy5eWMjOzkb79u2hra2NqKgoREdHiwlpVlaWWEd4eDhu3bqFo0ePIjQ0VOblivP27Vts374dAMQW65J8N6moqCA6Ohrr1q0DACgpKWHlypW4fv06tm3bhuPHjxfahxkZGVi8eDF27NiBkydP4sGDB1KfpYKOHz+Otm3bYu7cuZg0aZJM20REX6+8vDyEhITA0dER/v7+sLKygpqaGqysrODv7w9HR0fs3r1bpq7ipVkXEVGFICjQ77//LlSpUkVQU1MTnJ2dhSlTpghxcXHi/KioKEFHR0d49+6d1HK1a9cW1q9fLwiCIAQGBgoaGhpCamqqOH/ChAlCs2bNxPeurq7C6NGjpeqYPXu20K5dO6lpDx8+FAAIt27dEpf773//K1XGyclJmDRpkiAIgnDkyBFBSUlJLP8hX19fYejQoVLToqKiBCUlJeHt27dFLjNu3Dihc+fOgiAIwvLlywVPT0/B0dFRCAsLEwRBEKysrIQNGzYIgiAI9+7dEwAIly5dEgRBEE6cOCEAEI4dOybWd/DgQQGAuL7mzZsLI0aMkFpns2bNBEdHR/G9mZmZMHfu3ELbnb/clStXBIlEIjx//lx49eqVoKKiIsyePVvw9PQUBEEQ5syZIzg7Oxe5fYIgCMHBwQIA4fLly+K0+/fvC8rKysLjx4+lyrZp00aYMmWKuJyurm6h+mrUqCF0795datqH+0aW492tWzdh0KBB4vz169cLZmZmQm5u7ke3pW7dusL//vc/qViWLVv20fIf8vb2FgIDA2UuHxgYKCgpKQmamppSLw0NDantffz4saCsrCycO3dOEARByMrKEgwNDYWtW7fKFKuxsbEwfPhwQRD+b1+qq6sXWq+s8uuQ1fDhwwUdHZ1PluvXr5/Qtm1bqWkTJkwQ6tSpI76vUaOG8MMPP4jv8/LyhKpVqwpr164VBEEQduzYIdja2gp5eXlimczMTEFdXV04cuSIIAjvj5OxsbGQmZkplpFluQ99uC8lEokAQGjcuLGQlZUlCILs300NGzb85P757bffBAMDA/F9/mcvISFBnLZ69WrB2NhYfO/t7S1069ZN2Ldvn6ClpSWEhIQUu453794JKSkp4is/1pSUlE/GR2Xn3r17gre3txAdHS3cu3fvq3xFR0cL3t7ews2bu4WMjGOFXjdv7i52fmm98tezf/9+wdvbW7h3756iD2+puXHjhuDt7S3cuXOnyPl37twRvL29hRs3bpRpXURE5VVKSorMv3sUfo93586dERUVhbNnzyIsLAyLFi3Cpk2b4OPjg7i4OKSnpxe6j/Tt27e4e/eu+N7S0hLa2trie1NTUzx//rzYdcfFxeHEiRPQ0tIqNO/u3buwsbEBADg4OEjNK1j35cuXYW5uLpYtah1XrlzBzp07xWmCICAvLw/37t2Dvb19oWVcXV2xefNm5ObmIjIyEu3atYOJiQkiIiLg4OCAhIQEuLm5FbttBWM2NTUFADx//hzVq1dHfHw8/Pz8pMo3b95cbOlNTU3FkydP0KJFC6kyLVq0ELv416tXD/r6+oiMjISKigoaNmyILl26YPXq1QDet4B/KkYVFRWpOK9evYrc3NxC+zIzM1Om+4g/dZ+YLMfby8sLQ4YMwZo1a6CqqoqdO3eiT58+4r3n6enpCAoKwsGDB5GUlIScnBy8ffu2RC3eO3fuxLBhw6S2TyKRSHULDwsLQ8uWLT9ah62tLQ4cOCA17fHjx1L73MzMDJ07d8aWLVvQtGlT/PXXX8jMzESvXr1kilMQBEgkEqlpu3fvLvKc/Zi6devi/v37Yn0ApPZ/y5YtERYWJvP6ixIfH49u3bpJTWvRogWWL1+O3NxcsfdJwXNNIpHAxMRE/BzHxcUhISFB6jsEAN69eyf1PVO/fn2p+6hlXa4ou3fvhp2dHa5du4aJEydi69atYrd9Wb+bGjduXGj+sWPHMH/+fNy8eROpqanIycnBu3fvkJGRAQ0NDQCAhoYGateuLS5T1PfluXPnEBoait9///2TI5zPnz8fM2fOLLYMlR8f3mZBn2/fvn2KDqHUpaSkAADMzc2LnF+tWjWpcmVVFxFRRaDQxBsA1NTU0LZtW7Rt2xbTp0/H4MGDERgYCB8fH6Snp8PU1FSqW3O+gqNbf3ifqUQi+WTXpfT0dHTt2rXI+xbzk9VP1a2urv7JdQwbNgz+/v6F5lWvXr3IZVxcXJCWloaLFy/i5MmTmDdvHkxMTLBgwQI4OjrCzMwM1tbWxa63YMz5yUtpduWSSCRwcXFBREQEVFVV4ebmBgcHB2RmZuLatWs4ffr0R7uu5lNXV5dKrNLT06GsrIwLFy6IyVK+ohKQD2lqahY7X5bj3bVrVwiCgIMHD8LJyQlRUVFYtmyZWC4gIABHjx7F4sWLYWVlBXV1dfTs2VPmbsUA8N1336FZs2bi+0mTJqFatWpS50j+j5GPUVFRKdRdvFKlwh/lwYMHo3///li2bBmCg4Ph6ekpJl/FSU5OxosXL1CzZk2p6RYWFp/spl7QoUOHkJ2dDeD/LgwUvOe+uM+PjY0NUlJSkJSUJPV5/FzFfY7T09PRuHFjqQtk+YyMjMT/f3iOybpcUSwsLGBtbQ1ra2vk5OTAw8MD165dg6qqqszfTR/Gk5iYiC5dumD48OGYO3cu9PX1cerUKfj6+iIrK0s89kXtC+GDQe9q164NAwMDbNmyBZ07dy72Xv4pU6Zg3Lhx4vvU1FRxHAoqf4YOHQozMzNFh/FZnjx5Uq4uHHz//fcVLvnW1dUF8H6cjaK+7/OfKJFfrqzqIiKqCBSeeH+oTp06+PPPPwEAjRo1wtOnT1GpUiVYWlp+dp0qKirIzc2VmtaoUSNxpOSikhZZODg44NGjR7h9+3aRrd6NGjXCjRs3SpSs6OnpwcHBAatWrULlypVhZ2eHqlWrwtPTE6GhoZ+8v/tT7O3tce7cOQwYMECcVvCRUDo6OjAzM0N0dLTUuqKjo9G0aVPxvaurKzZu3AhVVVXMnTsXSkpKcHFxwc8//4zMzMxCLeaf0rBhQ+Tm5uL58+cfbe0t6jjKSpbjraamhu+//x47d+5EQkICbG1t0ahRI3F+dHQ0fHx84OHhAeB94iXLQFoFaWtrS7WQamtrQ19fv0TniKw6deoETU1NrF27FocPH8bJkydlWm7FihVQUlL64mc516hRQ/x//j6XdTt79uyJyZMnY9GiRVIXP/K9fv0aenp6sLe3L/T4wejoaNjY2BS6gPMxjRo1wu7du1G1alXo6OjItMyXLPehnj17YsaMGVizZg3Gjh372d9NFy5cQF5eHpYsWSL20tizZ89nxWRoaIh9+/bBzc0NvXv3xp49ez6afKuqqkJVVfWz1kNlz8zM7Iv+ntL/MTQ0VHQIpc7W1haGhoYIDQ2Fv7+/1NNG8vLyEBoaCiMjI3GAybKqi4ioIlDY4GrJyclo3bo1fvnlF1y5cgX37t3Db7/9hkWLFoldR93d3dG8eXN0794df//9NxITE3H69GlMnTq1RCNhWlpa4ty5c0hMTMTLly+Rl5eHkSNH4tWrV+jbty9iYmJw9+5dHDlyBAMHDpQ5uXN1dYWLiwt69OiBo0eP4t69ewgLCxNHAJ80aRJOnz6NUaNG4fLly7hz5w7279//0cHV8rm5uWHnzp1i4quvrw97e3vs3r37ixPv0aNHY8uWLQgODsbt27cRGBiI69evS5WZMGECFi5ciN27d+PWrVuYPHkyLl++jNGjR0vFeOPGDVy/fh3//e9/peJu0qTJJ1ugP5Tf1XvAgAHYt28f7t27h/Pnz2P+/Pk4ePAggPfHMT09HeHh4Xj58iUyMjJkrl/W4+3l5YWDBw9iy5Yt4qBq+aytrbFv3z5cvnwZcXFx6NevX7keFEZZWRk+Pj6YMmUKrK2t0bx580Jl0tLS8PTpUzx8+BAnT57E0KFDMWfOHMydO7dQkpycnIynT59KvT4cEb+0WFhYYNmyZVixYgV8fX0RGRmJ+/fvIzo6GsOGDcPs2bMBAOPHj0d4eDhmz56N27dvY9u2bVi1atUne1wU5OXlBUNDQ3Tr1g1RUVG4d+8eIiIi4O/vj0ePHpX6ch+SSCTw9/fHggULkJGR8dnfTVZWVsjOzsb//vc//PPPP9ixY4c46NrnqFq1Ko4fP46bN2+ib9++yMnJ+ey6iOjroKSkhD59+iAuLg4rV65EQkIC3r59i4SEBKxcuRJxcXHw9PSU6fGfpVkXEVFFoNBRzZs1a4Zly5bBxcUF9erVw/Tp0zFkyBCsWrUKwPsfpIcOHYKLiwsGDhwIGxsb9OnTB/fv3y/R824DAgKgrKyMOnXqwMjICA8ePBBbdXNzc9GuXTvUr18fY8aMgZ6eXon+COzduxdOTk7o27cv6tSpg4kTJ4o/jh0cHBAZGYnbt2+jZcuWaNiwIWbMmPHJbn6urq7Izc2VumfXzc2t0LTP4enpienTp2PixIlo3Lgx7t+/j+HDh0uV8ff3x7hx4zB+/HjUr18fhw8fxoEDB6S6uNevXx96enpo0KCB2BX8S2MMDg7GgAEDMH78eNja2qJ79+6IiYkRu+U7OzvDz88Pnp6eMDIywqJFi2SuW9bj3bp1a+jr6+PWrVvo16+fVB1Lly5FlSpV4OzsjK5du6J9+/ZSLeLlUX4344EDBxY5f8aMGTA1NYWVlRX69++PlJQUhIeHFzmCtbu7O0xNTaVe+b1T5GHEiBH4+++/8fjxY3h4eMDOzg6DBw+Gjo6OmFg3atQIe/bsQUhICOrVq4cZM2Zg1qxZ8PHxkXk9GhoaOHnyJKpXr47vv/8e9vb24mP2imvJ/tzliuLt7Y3s7GysWrXqs7+bHB0dsXTpUixcuBD16tXDzp07MX/+/BLF8SETExMcP34cV69ehZeX12f3OCGir0eTJk0wcuRIPHr0CHPmzMHw4cMxZ84cPH78GCNHjizRs7dLsy4ioq+dRPjw5j4iqjCioqLQpk0bPHz4sEQXq4g+V2pqKnR1dZGSkvJFXfCpdCUmJiIoKAhBQUFfbVfz/G2YMqUTqlcvPOjmgwfJmD//0Efnl5b89QwdOhQbNmz4qvdpcfLy8nDr1i2kpKRAV1cXtra2n906XZp1ERGVJyX53VPu7vEmoi+XmZmJFy9eICgoCL169WLSTUREJaKkpFSiJ1mUVV1ERF8rXm4kqoB+/fVX1KhRA69fvy5Rl3wiIiIiIip9TLyJKiAfHx/k5ubiwoULn3w8GRERERERyRcTbyIiogrO1NQUQUFBUs+Cpy9jaGjIfUpERDLjPd5EREQVnKqqaoUcAEyRVFRUuE+JiEhmbPEmIiIiIiIikiMm3kRERERERERyxMSbiIiIiIiISI6YeBMRERERERHJERNvIiIiIiIiIjniqOZERET01Xj6NKXY6R+bL+/1ExERFYeJNxEREZV72traUFFRQXBwdLHlPjW/NKioqEBbW1vu6yEioopDIgiCoOggiIioYkhNTYWuri5SUlKgo6Oj6HCogklOTkZaWpqiw4C2tjYMDAwUHQYRESlYSX73sMWbiIiIvgoGBgZMeImI6KvEwdWIiIiIiIiI5IiJNxEREREREZEcMfEmIiIiIiIikiMm3kRERERERERyxMSbiIiIiIiISI44qjkRERERlTvl5fFxJcFHzRHRxzDxJiIiIqJyJTk5GVOmTEFWVpaiQykRFRUVzJ8/n8k3ERXCxJuIiIiIypW0tDRkZWXBtWUX6OqWThL7OiUZJ6NC4dKyC/RKqc6CUlKSERkVirS0NCbeRFQIE28iIiIiKpd0dQ1gaGBSqnXqyaFOIqJP4eBqRERERERERHLExJuIiIiIiIhIjph4ExEREREREckRE28iIiIiIiIiOWLiTURERERERCRHTLyJiIiISK4yMzORmJiIzMxMRYfyVeD+Iqp4mHgTERERkVwlJSUhKCgISUlJig7lq8D9RVTxMPEmIiIiIiIikiMm3kRERERERERyxMSb6Cvi4+OD7t27f3E9EokEf/755xfX8zX4lra1NLi5uWHMmDGKDoOIiIioQmHiXUZ8fHwgkUggkUhQuXJlGBsbo23bttiyZQvy8vIUHV6Z2bp1K/T09BQdxldrxYoV2Lp1q6LDKDN79+6Fm5sbdHV1oaWlBQcHB8yaNQuvXr1SdGifJSIiQvwekEgkMDIyQqdOnXD16lVFh0ZERFRm8vLyEB8fj7NnzyI+Pv6b+i1M3y4m3mWoQ4cOSEpKQmJiIsLCwtCqVSuMHj0aXbp0QU5OjqLDozKUnZ39Wcvp6up+lRcuIiIiYGlpWaJlpk6dCk9PTzg5OSEsLAzXrl3DkiVLEBcXhx07dsgnUBllZWV90fK3bt1CUlISjhw5gszMTHTu3PmL6vzSeIiIiMpKbGwsJk6ciIULF2LdunVYuHAhJk6ciNjYWEWHRiRXTLzLkKqqKkxMTFCtWjU0atQIP/30E/bv34+wsDCpVswHDx6gW7du0NLSgo6ODnr37o1nz55J1fXXX3/ByckJampqMDQ0hIeHhzivqK61enp64joSExMhkUiwZ88etGzZEurq6nBycsLt27cRExODJk2aQEtLCx07dsSLFy+k6tm0aRPs7e2hpqYGOzs7rFmzRpyXX+++ffvQqlUraGhowNHREWfOnAHwPvkaOHAgUlJSxBa/oKCgIvdVXFwcWrVqBW1tbejo6KBx48biF3JQUBAaNGggVX758uVSiV1+l+x58+bB2NgYenp6mDVrFnJycjBhwgTo6+vD3NwcwcHBheIv6X6JiYlB27ZtYWhoCF1dXbi6uuLixYtS8UkkEqxduxbfffcdNDU1MWfOHFhZWWHx4sVS5S5fvgyJRIKEhIQi98uHXc3d3Nzg7++PiRMnQl9fHyYmJoX26Z07d+Di4gI1NTXUqVMHR48eLVTvw4cP0bt3b+jp6UFfXx/dunVDYmIiAODmzZvQ0NDArl27xPJ79uyBuro6bty4UWScX+r8+fOYN28elixZgp9//hnOzs6wtLRE27ZtsXfvXnh7e4tl165di9q1a0NFRQW2trZFJuUvX76Eh4cHNDQ0YG1tjQMHDkjNv3btGjp27AgtLS0YGxujf//+ePnypTjfzc0No0aNwpgxY2BoaIj27dvLtNzHVK1aFSYmJmjUqBHGjBmDhw8f4ubNm+L8U6dOieeghYUF/P398ebNG3G+paUlZs+ejQEDBkBHRwdDhw4FAEyaNAk2NjbQ0NBArVq1MH36dKmLPPmfnR07dsDS0hK6urro06cP0tLSPhrrwYMHoauri507d35yu4iIiIoTGxuL1atXw9zcHNOmTcO6deswbdo0mJubY/Xq1Uy+qUJj4q1grVu3hqOjI/bt2wfgfdebbt264dWrV4iMjMTRo0fxzz//wNPTU1zm4MGD8PDwQKdOnXDp0iWEh4ejadOmJV53YGAgpk2bhosXL6JSpUro168fJk6ciBUrViAqKgoJCQmYMWOGWH7nzp2YMWMG5s6di/j4eMybNw/Tp0/Htm3bpOqdOnUqAgICcPnyZdjY2KBv377IycmBs7Mzli9fDh0dHSQlJSEpKQkBAQFFxubl5QVzc3PExMTgwoULmDx5MipXrlyi7Tt+/DiePHmCkydPYunSpQgMDESXLl1QpUoVnDt3Dn5+fhg2bBgePXr0RfslLS0N3t7eOHXqFM6ePQtra2t06tSpUDITFBQEDw8PXL16Fb6+vhg0aJBU4g8AwcHBcHFxgZWVlczbuW3bNmhqauLcuXNYtGgRZs2aJSbXeXl5+P7776GiooJz585h3bp1mDRpktTy2dnZaN++PbS1tREVFYXo6GhoaWmhQ4cOyMrKgp2dHRYvXowRI0bgwYMHePToEfz8/LBw4ULUqVNH5jhLYufOndDS0sKIESOKnJ/f6v/HH39g9OjRGD9+PK5du4Zhw4Zh4MCBOHHihFT5mTNnonfv3rhy5Qo6deoELy8vsbv669ev0bp1azRs2BCxsbE4fPgwnj17ht69e0vVsW3bNqioqCA6Ohrr1q2TebnipKSkICQkBACgoqICALh79y46dOiAHj164MqVK9i9ezdOnTqFUaNGSS27ePFiODo64tKlS5g+fToAQFtbG1u3bsWNGzewYsUKbNy4EcuWLZNa7u7du/jzzz8RGhqK0NBQREZGYsGCBUXGt2vXLvTt2xc7d+6El5eXzNtFRET0oby8PISEhMDR0RH+/v6wsrKCmpoarKys4O/vD0dHR+zevZvdzqnCqqToAAiws7PDlStXAADh4eG4evUq7t27BwsLCwDA9u3bUbduXcTExMDJyQlz585Fnz59MHPmTLEOR0fHEq83ICBAbLkbPXo0+vbti/DwcLRo0QIA4OvrK9USHxgYiCVLluD7778HANSsWRM3btzA+vXrpVogAwIC0LlzZwDvE566desiISEBdnZ20NXVhUQigYmJSbGxPXjwABMmTICdnR0AwNrausTbp6+vj5UrV0JJSQm2trZYtGgRMjIy8NNPPwEApkyZggULFuDUqVPo06fPZ++X1q1bS613w4YN0NPTQ2RkJLp06SJO79evHwYOHCi+9/HxwYwZM3D+/Hk0bdoU2dnZ2LVrV6FW8E9xcHBAYGAggPf7adWqVQgPD0fbtm1x7Ngx3Lx5E0eOHIGZmRkAYN68eejYsaO4fP4fuU2bNkEikQB4fwFAT08PERERaNeuHUaMGIFDhw7hhx9+gIqKCpycnPDjjz+WKM6SuHPnDmrVqvXJiy2LFy+Gj4+PmKCPGzcOZ8+exeLFi9GqVSuxnI+PD/r27Qvg/favXLkS58+fR4cOHbBq1So0bNgQ8+bNE8tv2bIFFhYWuH37NmxsbAC837eLFi0Sy8yZM0em5Ypibm4OAGIr9nfffSee6/Pnz4eXl5c4wJm1tTVWrlwJV1dXrF27FmpqagDen3fjx4+XqnfatGni/y0tLREQEICQkBBMnDhRnJ6Xl4etW7dCW1sbANC/f3+Eh4dj7ty5UnWtXr0aU6dOxV9//QVXV9ePbktmZiYyMzPF96mpqR8tS0T05MmTUi1XHpVG7F/z9n/MrVu38PLlS/j5+UFJSbrtT0lJCV26dMGcOXNw69Yt2NvbKyhKIvlh4l0OCIIgJjzx8fGwsLAQk24AqFOnDvT09BAfHw8nJydcvnwZQ4YM+eL1Ojg4iP83NjYGANSvX19q2vPnzwG8TxDu3r0LX19fqXXn5ORAV1f3o/WampoCAJ4/fy4mFrIYN24cBg8ejB07dsDd3R29evVC7dq1S7B1QN26daW+2I2NjVGvXj3xvbKyMgwMDMRtLCr+T+0XAHj27BmmTZuGiIgIPH/+HLm5ucjIyMCDBw+k6m3SpInUezMzM3Tu3BlbtmxB06ZN8ddffyEzMxO9evUq0XYWjBd4v8/z48s/n/KTbgBo3ry5VPm4uDgkJCSIiVi+d+/e4e7du+L7LVu2wMbGBkpKSrh+/bp4zn6MlpaW+P/c3FxkZmZKTfvhhx+wbt26IpcVBKHYuvPFx8eL3azztWjRAitWrJCaVnAfaWpqQkdHR9xHcXFxOHHihFRs+e7evSsm0I0bN5aaJ+tyRYmKioKGhgbOnj2LefPmSe2HuLg4XLlyRaprtyAIyMvLw71798QfIx+eT8D7iygrV67E3bt3kZ6ejpycHOjo6EiVsbS0lDrWBc+XfL///jueP3+O6OhoODk5fXQ7gPcXCgpeBCQiKs6GDRsUHYLcfQvb+DlSUlIA/N/F5w9Vq1ZNqhxRRcPEuxyIj49HzZo1ZS6vrq5e7HyJRFIocSlqMK+CrYn5SdSH0/K7+6SnpwMANm7ciGbNmknVo6ys/Ml6S9ptKCgoCP369cPBgwcRFhaGwMBAhISEwMPDA0pKSiXevvxYipr2YWwl2S8A4O3tjeTkZKxYsQI1atSAqqoqmjdvXmjAK01NzUIxDh48GP3798eyZcsQHBwMT09PaGhoFLlPPkaWbSpOeno6GjduXOQ9vEZGRuL/4+Li8ObNGygpKSEpKUm8qPIxly9fFv9/7tw5TJo0CREREeK0DxPCgmxsbHDq1ClkZ2eX+BaDohS3j9LT09G1a1csXLiw0HIFt/HD4yfrckWpWbMm9PT0YGtri+fPn8PT0xMnT54U6x02bBj8/f0LLVe9evWPxnPmzBl4eXlh5syZaN++PXR1dRESEoIlS5ZIlZPlfGnYsCEuXryILVu2oEmTJsVeZJkyZQrGjRsnvk9NTZW6cEhEVNDQoUOlLgZ/zJMnT77aBFbWbSzO17z9H5PfUPPo0aMib6l7/PixVDmiioaJt4IdP34cV69exdixYwEA9vb2ePjwIR4+fCj+eL1x4wZev34t3k/r4OCA8PBwqW7LBRkZGSEpKUl8f+fOHWRkZHxRnMbGxjAzM8M///zzRfd6qqioIDc3V6ayNjY2sLGxwdixY9G3b18EBwfDw8MDRkZGePr0qVRPgYJJXlmLjo7GmjVr0KlTJwDvByqTZYAtAOjUqRM0NTWxdu1aHD58WEy+Skv++VQwUT579qxUmUaNGmH37t2oWrXqR5PhV69ewcfHB1OnTkVSUhK8vLxw8eLFYi8CFfyj+ujRI1SqVEnme9f79euHlStXYs2aNRg9enSh+a9fv4aenh7s7e0RHR0tdatDdHR0ie49b9SoEfbu3QtLS0tUqiT7V+LnLvehkSNHYv78+fjjjz/g4eGBRo0a4caNGyW6zx8ATp8+jRo1amDq1KnitPv3739WTLVr18aSJUvg5uYGZWVlrFq16qNlVVVVoaqq+lnrIaJvj5mZWYmfcvG1+Ra28XPY2trC0NAQoaGh8Pf3l+qVmJeXh9DQUBgZGcHW1laBURLJDwdXK0OZmZl4+vQpHj9+jIsXL2LevHno1q0bunTpggEDBgAA3N3dUb9+fTGxOX/+PAYMGABXV1exa2lgYCB+/fVXBAYGIj4+HlevXpVqdWvdujVWrVqFS5cuITY2Fn5+fqXSajhz5kzMnz8fK1euxO3bt3H16lUEBwdj6dKlMtdhaWmJ9PR0hIeH4+XLl0VeEHj79i1GjRqFiIgI3L9/H9HR0YiJiRG72Lq5ueHFixdYtGgR7t69i9WrVyMsLOyLt+9zWVtbY8eOHYiPj8e5c+fg5eX1yV4J+ZSVleHj44MpU6bA2tq6UDfwL+Xu7g4bGxt4e3sjLi4OUVFRUokZ8H4gO0NDQ3Tr1g1RUVG4d+8eIiIi4O/vLw485+fnBwsLC0ybNg1Lly5Fbm7uRwfGKw3NmjXDxIkTMX78eEycOBFnzpzB/fv3ER4ejl69eokD+k2YMAFbt27F2rVrcefOHSxduhT79u0rUWwjR47Eq1ev0LdvX8TExODu3bs4cuQIBg4cWOxFos9d7kMaGhoYMmQIAgMDIQgCJk2ahNOnT2PUqFG4fPky7ty5g/379xcaXO1D1tbWePDgAUJCQnD37l2sXLkSf/zxh8xxfMjGxgYnTpzA3r17xfvNiYiIPpeSkhL69OmDuLg4rFy5EgkJCXj79i0SEhKwcuVKxMXFwdPTs9D930QVBc/sMnT48GGYmprC0tISHTp0wIkTJ7By5Urs379f7K4tkUiwf/9+VKlSBS4uLnB3d0etWrWwe/dusR43Nzf89ttvOHDgABo0aIDWrVvj/Pnz4vwlS5bAwsICLVu2RL9+/RAQEFDi7stFGTx4MDZt2oTg4GDUr18frq6u2Lp1a4m6yTs7O8PPzw+enp4wMjKSGqwqn7KyMpKTkzFgwADY2Nigd+/e6Nixo3gfqb29PdasWYPVq1fD0dER58+fl2sS+CmbN2/Gv//+i0aNGqF///7w9/dH1apVZV7e19cXWVlZH+3B8CWUlJTwxx9/4O3bt2jatCkGDx5caBAtDQ0NnDx5EtWrV8f3338Pe3t7+Pr64t27d9DR0cH27dtx6NAh7NixA5UqVYKmpiZ++eUXbNy4Ua4XPBYuXIhdu3bh3LlzaN++PerWrYtx48bBwcFBbOHu3r07VqxYgcWLF6Nu3bpYv349goOD4ebmJvN6zMzMEB0djdzcXLRr1w7169fHmDFjoKenV+wf/89driijRo1CfHw8fvvtNzg4OCAyMhK3b99Gy5Yt0bBhQ8yYMeOT3Ra/++47jB07FqNGjUKDBg1w+vRpcbTzz2Vra4vjx4/j119/LTSQGxERUUk1adIEI0eOxKNHjzBnzhwMHz4cc+bMwePHjzFy5Mgixy8hqigkgqyjGBGRXERFRaFNmzZ4+PChOJgb0dcqNTUVurq6SElJKfY+fiL6tiQmJiIoKAhBQUEydcPOL/9dF28YGhT/JBRZvUx+igOh20q1zqLql3Ubi1PS/fW1ycvLw61bt5CSkgJdXV3Y2tqypZu+SiX53cN7vIkUJDMzEy9evEBQUBB69erFpJuIiIi+CUpKSnxkGH1zeGmJSEF+/fVX1KhRA69fvy6yyz0REREREVUMTLyJFMTHxwe5ubm4cOGC+OxKIiIiIiKqeJh4ExEREZFcmZqaIigoSHy0JRWP+4uo4uE93kREREQkV6qqqhVykDB54f4iqnjY4k1EREREREQkR0y8iYiIiIiIiOSIiTcRERERERGRHDHxJiIiIiIiIpIjJt5EREREREREcsRRzYmIiIioXEpJSS61ul7//7pel2KdBZVmrERU8TDxJiIiIqJyRVtbGyoqKoiMCi31uk/Koc58Kioq0NbWllv9RPT1kgiCICg6CCIiqhhSU1Ohq6uLlJQU6OjoKDocIvqKJScnIy0tTdFhlIi2tjYMDAwUHQYRlZGS/O5hizcRERERlTsGBgZMYomowuDgakRERERERERyxMSbiIiIiIiISI6YeBMRERERERHJERNvIiIiIiIiIjli4k1EREREREQkRxzVnIiIiIioFJXVo9D4+DKirwcTbyIiIiKiUpKcnIwpkycjKztb7utSqVwZ8xcsYPJN9BVg4k1EREREVErS0tKQlZ2NvtWNUVW18ifLP3+XhV8fPkdfi6qoqqYi83qeZ2bj1wfPkJaWxsSb6CvAxJuIiIiIqJRVVa0Mcw012curqZSoPBF9XTi4GhEREREREZEcMfEmIiIiIiIikiMm3kRERERERERyxMSbiIiIiIiISI6YeBMRERERERHJERNvIiIiIqJiZGZmIjExEZmZmYoORS4q+vYRlQdMvImIiIiIipGUlISgoCAkJSUpOhS5qOjbR1QeMPEmIiIiIiIikiMm3kRERERERERyxMSbvno+Pj7o3r37F9cjkUjw559/fnE9VH4EBQWhQYMGig7jqxEREQGJRILXr18rOhQiIiKiCkWhibePjw8kEgkkEgkqV64MY2NjtG3bFlu2bEFeXp4iQytTW7duhZ6enqLD+GqtWLECW7duVXQYcpeYmAiJRILLly8Xmufm5oYxY8bIXFdQUJD42atUqRIMDQ3h4uKC5cuXFxpYxc3NTSxb8OXn5/eFW1S8p0+f4scff0StWrWgqqoKCwsLdO3aFeHh4XJdrzwV3JdqamqwsbHB/PnzIQiCokMjIvoq5OXlIT4+HmfPnkV8fPw39XuRiL5ulRQdQIcOHRAcHIzc3Fw8e/YMhw8fxujRo/H777/jwIEDqFRJ4SFSGcnOzkblypVLvJyurq4copG/iIgI+Pj4IDExUSHrr1u3Lo4dO4a8vDwkJycjIiICc+bMwY4dOxAREQFtbW2x7JAhQzBr1iyp5TU0NGRel5ubG3x8fODj4yNT+cTERLRo0QJ6enr4+eefUb9+fWRnZ+PIkSMYOXIkbt68KfO6S5sgCMjNzf3s76b8fZmZmYnjx49j6NCh0NPTw/Dhwz+rvtzcXEgkEigpsQMTEVVssbGxCAkJwcuXL8VphoaG6NOnD5o0aaLAyIiIPk3hv9RUVVVhYmKCatWqoVGjRvjpp5+wf/9+hIWFSbViPnjwAN26dYOWlhZ0dHTQu3dvPHv2TKquv/76C05OTlBTU4OhoSE8PDzEeUV1I9bT0xPXkd+auGfPHrRs2RLq6upwcnLC7du3ERMTgyZNmkBLSwsdO3bEixcvpOrZtGkT7O3toaamBjs7O6xZs0acl1/vvn370KpVK2hoaMDR0RFnzpwB8D75GjhwIFJSUsSWsKCgoCL3VVxcHFq1agVtbW3o6OigcePGiI2NBVB0l9rly5fD0tJSfJ/fJXvevHkwNjaGnp4eZs2ahZycHEyYMAH6+vowNzdHcHBwofhLul9iYmLQtm1bGBoaQldXF66urrh48aJUfBKJBGvXrsV3330HTU1NzJkzB1ZWVli8eLFUucuXL0MikSAhIaHI/fJhV3M3Nzf4+/tj4sSJ0NfXh4mJSaF9eufOHbi4uEBNTQ116tTB0aNHC9X78OFD9O7dG3p6etDX10e3bt3EJPnmzZvQ0NDArl27xPJ79uyBuro6bty4UWScZWXWrFmoV69eoekNGjTA9OnTxfeVKlWCiYkJzMzMUL9+ffz444+IjIzEtWvXsHDhQqllNTQ0YGJiIvXS0dGR2zaMGDECEokE58+fR48ePWBjY4O6deti3LhxOHv2rFhOlu8FANixYwcsLS2hq6uLPn36IC0tTZyXl5eH+fPno2bNmlBXV4ejoyN+//13cX5+9+uwsDA0btwYqqqqOHXq1CeX+5j8fVmjRg0MHDgQDg4OUudfZmYmAgICUK1aNWhqaqJZs2aIiIgQ5+f3kDlw4ADq1KkDVVVVPHjwQObP3KZNm+Dh4QENDQ1YW1vjwIEDH401IyMDHTt2RIsWLdj9nIgUKjY2FqtXr4a5uTmmTZuGdevWYdq0aTA3N8fq1avF30NEROWVwhPvorRu3RqOjo7Yt28fgPc/jLt164ZXr14hMjISR48exT///ANPT09xmYMHD8LDwwOdOnXCpUuXEB4ejqZNm5Z43YGBgZg2bRouXryISpUqoV+/fpg4cSJWrFiBqKgoJCQkYMaMGWL5nTt3YsaMGZg7dy7i4+Mxb948TJ8+Hdu2bZOqd+rUqQgICMDly5dhY2ODvn37IicnB87Ozli+fDl0dHSQlJSEpKQkBAQEFBmbl5cXzM3NERMTgwsXLmDy5MklbiE+fvw4njx5gpMnT2Lp0qUIDAxEly5dUKVKFZw7dw5+fn4YNmwYHj169EX7JS0tDd7e3jh16hTOnj0La2trdOrUSSrhAd5fMPDw8MDVq1fh6+uLQYMGSSX+ABAcHAwXFxdYWVnJvJ3btm2DpqYmzp07h0WLFmHWrFlicpOXl4fvv/8eKioqOHfuHNatW4dJkyZJLZ+dnY327dtDW1sbUVFRiI6OhpaWFjp06ICsrCzY2dlh8eLFGDFiBB48eIBHjx7Bz88PCxcuRJ06dWSOUx4GDRqE+Ph4xMTEiNMuXbqEK1euYODAgcUua2dnh44dO4qfPUV49eoVDh8+jJEjR0JTU7PQ/PzbMmT5XgCAu3fv4s8//0RoaChCQ0MRGRmJBQsWiPPnz5+P7du3Y926dbh+/TrGjh2LH374AZGRkVL1TJ48GQsWLEB8fDwcHBxkXu5jBEFAVFQUbt68CRUVFXH6qFGjcObMGYSEhODKlSvo1asXOnTogDt37ohlMjIysHDhQmzatAnXr19H1apVZf7MzZw5E71798aVK1fQqVMneHl54dWrV4Xie/36Ndq2bYu8vDwcPXqUt8MQkcLk5eUhJCQEjo6O8Pf3h5WVFdTU1GBlZQV/f384Ojpi9+7d7HZOROVaue3HbWdnhytXrgAAwsPDcfXqVdy7dw8WFhYAgO3bt6Nu3bqIiYmBk5MT5s6diz59+mDmzJliHY6OjiVeb0BAANq3bw8AGD16NPr27Yvw8HC0aNECAODr6yvVEh8YGIglS5bg+++/BwDUrFkTN27cwPr16+Ht7S1Vb+fOnQG8/+Fbt25dJCQkwM7ODrq6upBIJDAxMSk2tgcPHmDChAmws7MDAFhbW5d4+/T19bFy5UooKSnB1tYWixYtQkZGBn766ScAwJQpU7BgwQKcOnUKffr0+ez90rp1a6n1btiwAXp6eoiMjESXLl3E6f369ZNKBn18fDBjxgycP38eTZs2RXZ2Nnbt2lWoFfxTHBwcEBgYCOD9flq1ahXCw8PRtm1bHDt2DDdv3sSRI0dgZmYGAJg3bx46duwoLp//B3zTpk2QSCQA3l8A0NPTQ0REBNq1a4cRI0bg0KFD+OGHH6CiogInJyf8+OOPJYrzczg7OxfqVvz27Vuxx4O5uTnat2+P4OBgODk5ibG7urqiVq1an6zfzs4Of//9t9S0NWvWYNOmTVLT1q9fDy8vry/YkqIlJCRAEATxPP8YWb4XgPc/2LZu3Sp2ne/fvz/Cw8Mxd+5cZGZmYt68eTh27BiaN28OAKhVqxZOnTqF9evXw9XVVVzfrFmz0LZtWwAo0XIfyt+XWVlZyM7OhpqaGvz9/QG8/4wHBwfjwYMH4rkZEBCAw4cPIzg4GPPmzQPw/sLQmjVrpL7jZP3M+fj4oG/fvgDen/crV67E+fPn0aFDB7HM06dP4enpCWtra+zatUvqwsCHMjMzpcYFSE1N/WhZIqLPcevWLbx8+RJ+fn6F/v4pKSmhS5cumDNnDm7dugV7e3u5xvLkyZNSKVOaSmN9ZR0z0beo3CbegiCICU98fDwsLCzEH9cAUKdOHejp6SE+Ph5OTk64fPkyhgwZ8sXrdXBwEP9vbGwMAKhfv77UtOfPnwMA3rx5g7t378LX11dq3Tk5OYXuOy5Yr6mpKQDg+fPnn0wuCho3bhwGDx6MHTt2wN3dHb169ULt2rVLsHXv7+st+EfL2NhYqluysrIyDAwMxG0sKv5P7RcAePbsGaZNm4aIiAg8f/4cubm5yMjIwIMHD6Tq/fCeLDMzM3Tu3BlbtmxB06ZN8ddffyEzMxO9evUq0XYWjBd4v8/z48s/n/ITGwBi8pQvLi4OCQkJUvc5A8C7d+9w9+5d8f2WLVtgY2MDJSUlXL9+XTxnP0ZLS0v8f25uLjIzM6Wm/fDDD1i3bl2xdezevbvQD4sPE+AhQ4Zg0KBBWLp0KZSUlLBr1y4sW7as2HrzFfzsFax/6tSpUtPyz4OizJs3T0wSgfcXBs6ePYtRo0aJ027cuIHq1asXuX5ZyPK9AACWlpZSx7HguZCQkICMjAwxoc6XlZWFhg0bSk0reK6WZLkP5e/Lf//9F4GBgXB2doazszMA4OrVq8jNzYWNjY3UMpmZmTAwMBDfq6ioFDrHZf3MFVxOU1MTOjo6hT7vbdu2RdOmTbF7924oKysXuz3z58+XuuBJRFTaUlJSALy/sFyUatWqSZWTpw0bNsh9HSVVHmMiosLKbeIdHx+PmjVrylxeXV292PkSiaTQD/rs7OxC5Qp23c5PPj6clt+VKT09HQCwceNGNGvWTKqeD3+sFlVvSbtEBQUFoV+/fjh48CDCwsIQGBiIkJAQeHh4QElJqcTblx9LUdM+jK0k+wUAvL29kZycjBUrVqBGjRpQVVVF8+bNkZWVJVVvUV2JBw8ejP79+2PZsmUIDg6Gp6dniQby+th2lmR/p6eno3Hjxti5c2eheUZGRuL/4+Li8ObNGygpKSEpKUm8qPIxBUckP3fuHCZNmiR1/64s901bWFgU6nb/4fnftWtXqKqq4o8//oCKigqys7PRs2fPT9YNFP3Z09XVLVFXfz8/P/Tu3Vt87+XlhR49eog9QwBIXfgoyNraGhKJpNQGUCvuXMj/DB88eFD84ZZPVVVV6n3Bc7Uky32o4L7cs2cPrKys8J///Afu7u5IT0+HsrIyLly4UOg7pOAFGnV19UIXR2T9zMny2ejcuTP27t2LGzduSF1gK8qUKVMwbtw48X1qaqrUxRAioi+V35jx6NGjIv8WPX78WKqcPA0dOvSjf7/yPXnypEyTYVli+pSyjpnoW1QuE+/jx4/j6tWrGDt2LADA3t4eDx8+xMOHD8UfdDdu3MDr16/F+2kdHBwQHh7+0XtYjYyMkJSUJL6/c+cOMjIyvihOY2NjmJmZ4Z9//vmiLrcqKirIzc2VqayNjQ1sbGwwduxY9O3bF8HBwfDw8ICRkRGePn0q1VpZ1GOnykp0dDTWrFmDTp06AXg/UFnBUUiL06lTJ2hqamLt2rU4fPgwTp48Waqx5Z9PBRPlggN2AUCjRo2we/duVK1a9aPJ8KtXr+Dj44OpU6ciKSkJXl5euHjxYrEXgQr+YHj06BEqVapUooRWVpUqVYK3tzeCg4OhoqKCPn36fPLiFPB+0LjDhw9jypQpX7R+fX196Ovri+/V1dVRtWpVmbZVX18f7du3x+rVq+Hv71/o4szr16+hp6cn0/fCpxQcnKy47uGltdyHtLS0MHr0aAQEBODSpUto2LAhcnNz8fz5c7Rs2bJEdX3JZ+5DCxYsgJaWFtq0aYOIiIhi96eqquonLzYQEX0JW1tbGBoaIjQ0FP7+/lI99/Ly8hAaGgojIyPY2trKPRYzMzOpgWvLg/IYExEVpvDEOzMzE0+fPpV6nNj8+fPRpUsXDBgwAADg7u6O+vXrw8vLC8uXL0dOTg5GjBgBV1dXsftnYGAg2rRpg9q1a6NPnz7IycnBoUOHxEGzWrdujVWrVqF58+bIzc3FpEmTPuvRVR+aOXMm/P39oauriw4dOiAzMxOxsbH4999/pVqBimNpaYn09HSEh4fD0dERGhoahVp43759iwkTJqBnz56oWbMmHj16hJiYGPTo0QPA+5G8X7x4gUWLFqFnz544fPgwwsLC5DrydHGsra2xY8cONGnSBKmpqZgwYYJMiR/wvreAj48PpkyZAmtr60LdwL+Uu7s7bGxs4O3tjZ9//hmpqamFulF7eXnh559/Rrdu3TBr1iyYm5vj/v372LdvHyZOnAhzc3P4+fnBwsIC06ZNQ2ZmJho2bIiAgACsXr26VOP9XIMHDxa7pEdHRxean5OTg6dPnxZ6nFiDBg0wYcIEqbIZGRl4+vSp1DRVVVVUqVJFLrGvXr0aLVq0QNOmTTFr1iw4ODggJycHR48exdq1axEfHy/T98KnaGtrIyAgAGPHjkVeXh7++9//IiUlBdHR0dDR0ZEap6E0livKsGHDMHv2bOzduxc9e/aEl5cXBgwYgCVLlqBhw4Z48eIFwsPD4eDgII4TUZQv+cwVZfHixcjNzUXr1q0RERFRottiiIhKk5KSEvr06YPVq1dj5cqV6NKlC6pVq4bHjx8jNDQUcXFxGDlyJB+rSETlmsK/oQ4fPgxTU1NYWlqiQ4cOOHHiBFauXIn9+/eLXS0lEgn279+PKlWqwMXFBe7u7qhVqxZ2794t1uPm5obffvsNBw4cQIMGDdC6dWucP39enL9kyRJYWFigZcuW6NevHwICAkrcfbkogwcPxqZNmxAcHIz69evD1dUVW7duLVE3eWdnZ/j5+cHT0xNGRkZYtGhRoTLKyspITk7GgAEDYGNjg969e6Njx47ivZX29vZYs2YNVq9eDUdHR5w/f/6jo6OXhc2bN+Pff/9Fo0aN0L9/f/j7+6Nq1aoyL+/r64usrKxPjsL9OZSUlPDHH3/g7du3aNq0KQYPHoy5c+dKldHQ0MDJkydRvXp1fP/997C3t4evry/evXsHHR0dbN++HYcOHcKOHTtQqVIlaGpq4pdffsHGjRsRFhZW6jF/Dmtrazg7O8POzq7QrRAAcP36dZiamqJ69epwc3PDnj17MGXKFERFRUl1awbe305hamoq9cofoEseatWqhYsXL6JVq1YYP3486tWrh7Zt2yI8PBxr164FINv3gixmz56N6dOnY/78+bC3t0eHDh1w8ODBT36GP3e5D+nr62PAgAEICgpCXl4egoODMWDAAIwfPx62trbo3r07YmJiirwfvqAv/cwVZdmyZejduzdat26N27dvf1FdRERfokmTJhg5ciQePXqEOXPmYPjw4ZgzZw4eP36MkSNH8jneRFTuSQRZRzIiKkNRUVFo06YNHj58WOwgXvRxgiDA2toaI0aMkLn3BdGXSk1Nha6uLlJSUhTW44aIKq68vDzcunULKSkp0NXVha2tbZm0dCcmJiIoKAhBQUGf7NadX3a0tTnMNdQ+WfejjHdYceeRzOU/XE6WmD6lJNtHRP+nJL97FN7VnKigzMxMvHjxAkFBQejVqxeT7s/04sULhISE4OnTp3LpNUBERKQISkpKcn9kGBGRPDDxpnLl119/ha+vLxo0aIDt27crOpyvVtWqVWFoaIgNGzbI7T5sIiIiIiKSDRNvKld8fHzg4+Oj6DC+eryDhIiIiIio/FD44GpEREREREREFRkTbyIiIiKiYpiamiIoKAimpqaKDkUuKvr2EZUH7GpORERERFQMVVXVCj3ad0XfPqLygC3eRERERERERHLExJuIiIiIiIhIjph4ExEREREREckRE28iIiIiIiIiOeLgakREREREpex5ZrZs5d5lSf1b2vUTUfnAxJuIiIiIqJRoa2tDpXJl/PrgWYmW+/Xh8xKvS6VyZWhra5d4OSIqexJBEARFB0FERBVDamoqdHV1kZKSAh0dHUWHQ0SkEMnJyUhLS5P7erS1tWFgYCD39RBR0Uryu4ct3kREREREpcjAwIAJMRFJ4eBqRERERERERHLEFm8iIio1+XcvpaamKjgSIiIiIvnK/70jy93bTLyJiKjU5N/TaGFhoeBIiIiIiMpGWloadHV1iy3DwdWIiKjU5OXl4cmTJ9DW1oZEIlF0OOVCamoqLCws8PDhQw44V47wuJRPPC7lD49J+cTjUj4IgoC0tDSYmZlBSan4u7jZ4k1ERKVGSUkJ5ubmig6jXNLR0eGPo3KIx6V84nEpf3hMyiceF8X7VEt3Pg6uRkRERERERCRHTLyJiIiIiIiI5IiJNxERkRypqqoiMDAQqqqqig6FCuBxKZ94XMofHpPyicfl68PB1YiIiIiIiIjkiC3eRERERERERHLExJuIiIiIiIhIjph4ExEREREREckRE28iIqLPNH/+fDg5OUFbWxtVq1ZF9+7dcevWrU8ut3z5ctja2kJdXR0WFhYYO3Ys3r17VwYRfxs+57hkZ2dj1qxZqF27NtTU1ODo6IjDhw+XUcTfhrVr18LBwUF87nDz5s0RFhZW7DK//fYb7OzsoKamhvr16+PQoUNlFO23oaTH5Pr16+jRowcsLS0hkUiwfPnysgv2G1LS47Jx40a0bNkSVapUQZUqVeDu7o7z58+XYcQkCybeREREnykyMhIjR47E2bNncfToUWRnZ6Ndu3Z48+bNR5fZtWsXJk+ejMDAQMTHx2Pz5s3YvXs3fvrppzKMvGL7nOMybdo0rF+/Hv/73/9w48YN+Pn5wcPDA5cuXSrDyCs2c3NzLFiwABcuXEBsbCxat26Nbt264fr160WWP336NPr27QtfX19cunQJ3bt3R/fu3XHt2rUyjrziKukxycjIQK1atbBgwQKYmJiUcbTfjpIel4iICPTt2xcnTpzAmTNnYGFhgXbt2uHx48dlHDkVh6OaExERlZIXL16gatWqiIyMhIuLS5FlRo0ahfj4eISHh4vTxo8fj3PnzuHUqVNlFeo3RZbjYmZmhqlTp2LkyJHitB49ekBdXR2//PJLWYX6zdHX18fPP/8MX1/fQvM8PT3x5s0bhIaGitP+85//oEGDBli3bl1ZhvlNKe6YFGRpaYkxY8ZgzJgxZRPYN07W4wIAubm5qFKlClatWoUBAwaUQXQkC7Z4ExERlZKUlBQA738gfYyzszMuXLggdgP8559/cOjQIXTq1KlMYvwWyXJcMjMzoaamJjVNXV2dF0PkJDc3FyEhIXjz5g2aN29eZJkzZ87A3d1dalr79u1x5syZsgjxmyPLMaGy9znHJSMjA9nZ2cV+51HZq6ToAIiIiCqCvLw8jBkzBi1atEC9evU+Wq5fv354+fIl/vvf/0IQBOTk5MDPz49dzeVE1uPSvn17LF26FC4uLqhduzbCw8Oxb98+5ObmlmG0Fd/Vq1fRvHlzvHv3DlpaWvjjjz9Qp06dIss+ffoUxsbGUtOMjY3x9OnTsgj1m1GSY0Jl50uOy6RJk2BmZlbowhUpFlu8iYiISsHIkSNx7do1hISEFFsuIiIC8+bNw5o1a3Dx4kXs27cPBw8exOzZs8so0m+LrMdlxYoVsLa2hp2dHVRUVDBq1CgMHDgQSkr8qVSabG1tcfnyZZw7dw7Dhw+Ht7c3bty4oeiwvmk8JuXT5x6XBQsWICQkBH/88UehXjykWGzxJiIi+kKjRo1CaGgoTp48CXNz82LLTp8+Hf3798fgwYMBAPXr18ebN28wdOhQTJ06lYleKSrJcTEyMsKff/6Jd+/eITk5GWZmZpg8eTJq1apVRtF+G1RUVGBlZQUAaNy4MWJiYrBixQqsX7++UFkTExM8e/ZMatqzZ884qFcpK8kxobLzOcdl8eLFWLBgAY4dOwYHB4eyCpVkxL/uREREn0kQBIwaNQp//PEHjh8/jpo1a35ymYyMjELJtbKyslgffbnPOS751NTUUK1aNeTk5GDv3r3o1q2bHCOlvLw8ZGZmFjmvefPmUoMQAsDRo0d5/7GcFXdMSHE+dVwWLVqE2bNn4/Dhw2jSpEkZRkayYos3ERHRZxo5ciR27dqF/fv3Q1tbW7z3VFdXF+rq6gCAAQMGoFq1apg/fz4AoGvXrli6dCkaNmyIZs2aISEhAdOnT0fXrl3FBJy+zOccl3PnzuHx48do0KABHj9+jKCgIOTl5WHixIkK246KZsqUKejYsSOqV6+OtLQ07Nq1CxEREThy5AiAwsdk9OjRcHV1xZIlS9C5c2eEhIQgNjYWGzZsUORmVCglPSZZWVlid+esrCw8fvwYly9fhpaWltg6S1+upMdl4cKFmDFjBnbt2gVLS0vxO09LSwtaWloK2w76gEBERESfBUCRr+DgYLGMq6ur4O3tLb7Pzs4WgoKChNq1awtqamqChYWFMGLECOHff/8t8/grqs85LhEREYK9vb2gqqoqGBgYCP379xceP35c9sFXYIMGDRJq1KghqKioCEZGRkKbNm2Ev//+W5z/4TERBEHYs2ePYGNjI6ioqAh169YVDh48WMZRV2wlPSb37t0r8rPl6upa9sFXYCU9LjVq1CjyuAQGBpZ98PRRfI43ERERERERkRzxHm8iIiIiIiIiOWLiTURERERERCRHTLyJiIiIiIiI5IiJNxEREREREZEcMfEmIiIiIiIikiMm3kRERERERERyxMSbiIiIiIiISI6YeBMRERERERHJERNvIiIiIqpwfHx80L179y+qIyIiAhKJBK9fvwYAbN26FXp6el8cW2JiIiQSCS5fvvzFdRHR14GJNxEREVEFdubMGSgrK6Nz586KDqVUbdy4EY6OjtDS0oKenh4aNmyI+fPni/NXrFiBrVu3ftE6nJ2dkZSUBF1d3S+MVpqFhQWSkpJQr149AIUTfCKqeCopOgAiIiIikp/Nmzfjxx9/xObNm/HkyROYmZkpOiSZZWVlQUVFpdD0LVu2YMyYMVi5ciVcXV2RmZmJK1eu4Nq1a2KZ0kiWVVRUYGJi8sX1FJS/TaVdLxGVb2zxJiIiIqqg0tPTsXv3bgwfPhydO3cu1AKc39IaHh6OJk2aQENDA87Ozrh165ZYJi4uDq1atYK2tjZ0dHTQuHFjxMbGQhAEGBkZ4ffffxfLNmjQAKampuL7U6dOQVVVFRkZGQCA169fY/DgwTAyMoKOjg5at26NuLg4sXxQUBAaNGiATZs2oWbNmlBTUytyuw4cOIDevXvD19cXVlZWqFu3Lvr27Yu5c+eKZT7sau7m5oYff/wRY8aMQZUqVWBsbIyNGzfizZs3GDhwILS1tWFlZYWwsLBC++djLdF3795Ft27dYGxsDC0tLTg5OeHYsWNSZSwtLTF79mwMGDAAOjo6GDp0qFRX88TERLRq1QoAUKVKFUgkEvj4+GD79u0wMDBAZmamVH3du3dH//79i4yHiMovJt5EREREFdSePXtgZ2cHW1tb/PDDD9iyZQsEQShUburUqViyZAliY2NRqVIlDBo0SJzn5eUFc3NzxMTE4MKFC5g8eTIqV64MiUQCFxcXREREAAD+/fdfxMfH4+3bt7h58yYAIDIyEk5OTtDQ0AAA9OrVC8+fP0dYWBguXLiARo0aoU2bNnj16pW4voSEBOzduxf79u376D3QJiYmOHv2LO7fv1+i/bFt2zYYGhri/Pnz+PHHHzF8+HD06tULzs7OuHjxItq1a4f+/fuLFwo+JT09HZ06dUJ4eDguXbqEDh06oGvXrnjw4IFUucWLF8PR0RGXLl3C9OnTpeZZWFhg7969AIBbt24hKSkJK1asQK9evZCbm4sDBw6IZZ8/f46DBw9KHR8i+jow8SYiIiKqoDZv3owffvgBANChQwekpKQgMjKyULm5c+fC1dUVderUweTJk3H69Gm8e/cOAPDgwQO4u7vDzs4O1tbW6NWrFxwdHQG8b0XOT7xPnjyJhg0bSk2LiIiAq6srgPet3+fPn8dvv/2GJk2awNraGosXL4aenp5Uq3lWVha2b9+Ohg0bwsHBocjtCgwMhJ6eHiwtLWFrawsfHx/s2bMHeXl5xe4PR0dHTJs2DdbW1pgyZQrU1NRgaGiIIUOGwNraGjNmzEBycjKuXLki0/51dHTEsGHDUK9ePVhbW2P27NmoXbu2VLIMAK1bt8b48eNRu3Zt1K5dW2qesrIy9PX1AQBVq1aFiYkJdHV1oa6ujn79+iE4OFgs+8svv6B69epwc3OTKT4iKj+YeBMRERFVQLdu3cL58+fRt29fAEClSpXg6emJzZs3FypbMMHN7yr+/PlzAMC4ceMwePBguLu7Y8GCBbh7965Y1tXVFTdu3MCLFy8QGRkJNzc3MfHOzs7G6dOnxSQxLi4O6enpMDAwgJaWlvi6d++eVJ01atSAkZFRsdtmamqKM2fO4OrVqxg9ejRycnLg7e2NDh06FJt8F9xOZWVlGBgYoH79+uI0Y2NjqW3/lPT0dAQEBMDe3h56enrQ0tJCfHx8oRbvJk2ayFTfh4YMGYK///4bjx8/BvB+VHUfHx9IJJLPqo+IFIeDqxERERFVQJs3b0ZOTo7UYGqCIEBVVRWrVq2SGnyscuXK4v/zk7r8BDYoKAj9+vXDwYMHERYWhsDAQISEhMDDwwP169eHvr4+IiMjERkZiblz58LExAQLFy5ETEwMsrOz4ezsDOB9kmpqaiq2hhdU8BFdmpqaMm9jvXr1UK9ePYwYMQJ+fn5o2bIlIiMjxXumP1RwO/O3tbht/5SAgAAcPXoUixcvhpWVFdTV1dGzZ09kZWVJlSvJNhXUsGFDODo6Yvv27WjXrh2uX7+OgwcPflZdRKRYTLyJiIiIKpicnBxs374dS5YsQbt27aTmde/eHb/++iv8/Pxkrs/GxgY2NjYYO3Ys+vbti+DgYHh4eEAikaBly5bYv38/rl+/jv/+97/Q0NBAZmYm1q9fjyZNmohJZ6NGjfD06VNUqlQJlpaWpbm5AIA6deoAAN68eVPqdX9MdHQ0fHx84OHhAeD9xYXExMQS15M/cntubm6heYMHD8by5cvx+PFjuLu7w8LC4otiJiLFYFdzIiIiogomNDQU//77L3x9fcVW4fxXjx49iuxuXpS3b99i1KhRiIiIwP379xEdHY2YmBjY29uLZdzc3PDrr7+iQYMG0NLSgpKSElxcXLBz507x/m4AcHd3R/PmzdG9e3f8/fffSExMxOnTpzF16lTExsaWaPuGDx+O2bNnIzo6Gvfv38fZs2cxYMAAGBkZoXnz5iWq60tYW1uLg8DFxcWhX79+MreWF1SjRg1IJBKEhobixYsXSE9PF+f169cPjx49wsaNGzmoGtFXjIk3ERERUQWzefNmuLu7F/ks6x49eiA2NlamAcSUlZWRnJyMAQMGwMbGBr1790bHjh0xc+ZMsYyrqytyc3OlBvxyc3MrNE0ikeDQoUNwcXHBwIEDYWNjgz59+uD+/fvivdWycnd3x9mzZ9GrVy/Y2NigR48eUFNTQ3h4OAwMDEpU15dYunQpqlSpAmdnZ3Tt2hXt27dHo0aNSlxPtWrVMHPmTEyePBnGxsYYNWqUOE9XVxc9evSAlpaW1OPRiOjrIhGKeqYEERERERGVC23atEHdunWxcuVKRYdCRJ+JiTcRERERUTn077//IiIiAj179sSNGzdga2ur6JCI6DNxcDUiIiIionKoYcOG+Pfff7Fw4UIm3URfObZ4ExEREREREckRB1cjIiIiIiIikiMm3kRERERERERyxMSbiIiIiIiISI6YeBMRERERERHJERNvIiIiIiIiIjli4k1EREREREQkR0y8iYiIiIiIiOSIiTcRERERERGRHDHxJiIiIiIiIpKj/wfEOMCB1qRWEgAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Statistical Analysis\n",
        "\n",
        "To statistically evaluate the differences in `retrieval_precision` and `answer_similarity` across various RAG techniques, we perform ANOVA and Tukey's Honestly Significant Difference (HSD) tests. These tests help us determine if the observed differences in metrics are statistically significant, providing a robust basis for comparing the performance of each technique.\n"
      ],
      "metadata": {
        "id": "2bUYLC6s8xCl"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### ANOVA"
      ],
      "metadata": {
        "id": "RUciEb1V8yVI"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from scipy.stats import f_oneway\n",
        "\n",
        "# ANOVA for retrieval precision and answer similarity\n",
        "groups_rp = expanded_data.groupby('Experiment')['retrieval_precision'].apply(list)\n",
        "groups_as = expanded_data.groupby('Experiment')['answer_similarity'].apply(list)\n",
        "\n",
        "anova_rp = f_oneway(*groups_rp)\n",
        "anova_as = f_oneway(*groups_as)\n",
        "\n",
        "anova_results = {\n",
        "    'Retrieval Precision': {'statistic': anova_rp.statistic, 'p-value': anova_rp.pvalue},\n",
        "    'Answer Similarity': {'statistic': anova_as.statistic, 'p-value': anova_as.pvalue}\n",
        "}\n",
        "\n",
        "anova_results\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "h08LkzsEK6N1",
        "outputId": "ad8404f2-054a-4ad5-e15e-0a7b21831663"
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'Retrieval Precision': {'statistic': 338.4387380531311,\n",
              "  'p-value': 3.903821221672523e-104},\n",
              " 'Answer Similarity': {'statistic': 77.77379495798307,\n",
              "  'p-value': 3.6586443584180474e-61}}"
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Tukey - all"
      ],
      "metadata": {
        "id": "fH4YnlGc9J2m"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from statsmodels.stats.multicomp import pairwise_tukeyhsd\n",
        "import pandas as pd\n",
        "\n",
        "tukey_rp = pairwise_tukeyhsd(endog=expanded_data['retrieval_precision'], groups=expanded_data['Experiment'], alpha=0.05)\n",
        "tukey_result_df = pd.DataFrame(data=tukey_rp.summary().data[1:], columns=tukey_rp.summary().data[0])\n",
        "tukey_result_df\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 423
        },
        "id": "3E1zbDldTupg",
        "outputId": "b35b565e-e8b4-497a-df73-75360a0c332d"
      },
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                               group1  \\\n",
              "0                         Classic VDB + Cohere Rerank   \n",
              "1                         Classic VDB + Cohere Rerank   \n",
              "2                         Classic VDB + Cohere Rerank   \n",
              "3                         Classic VDB + Cohere Rerank   \n",
              "4                         Classic VDB + Cohere Rerank   \n",
              "..                                                ...   \n",
              "115         Sentence window retrieval + Cohere rerank   \n",
              "116         Sentence window retrieval + Cohere rerank   \n",
              "117                  Sentence window retrieval + HyDE   \n",
              "118                  Sentence window retrieval + HyDE   \n",
              "119  Sentence window retrieval + HyDE + Cohere Rerank   \n",
              "\n",
              "                                               group2  meandiff   p-adj  \\\n",
              "0                                  Classic VDB + HyDE    0.0498  0.0000   \n",
              "1                  Classic VDB + HyDE + Cohere Rerank    0.0221  0.0219   \n",
              "2                     Classic VDB + HyDE + LLM Rerank    0.0599  0.0000   \n",
              "3                            Classic VDB + LLM Rerank    0.0364  0.0000   \n",
              "4      Classic VDB + Maximal Marginal Relevance (MMR)    0.0006  1.0000   \n",
              "..                                                ...       ...     ...   \n",
              "115  Sentence window retrieval + HyDE + Cohere Rerank   -0.0168  0.2521   \n",
              "116            Sentence window retrieval + LLM Rerank    0.0241  0.0067   \n",
              "117  Sentence window retrieval + HyDE + Cohere Rerank   -0.0053  0.9999   \n",
              "118            Sentence window retrieval + LLM Rerank    0.0357  0.0000   \n",
              "119            Sentence window retrieval + LLM Rerank    0.0410  0.0000   \n",
              "\n",
              "      lower   upper  reject  \n",
              "0    0.0293  0.0704    True  \n",
              "1    0.0016  0.0427    True  \n",
              "2    0.0393  0.0805    True  \n",
              "3    0.0159  0.0570    True  \n",
              "4   -0.0199  0.0212   False  \n",
              "..      ...     ...     ...  \n",
              "115 -0.0374  0.0037   False  \n",
              "116  0.0036  0.0447    True  \n",
              "117 -0.0259  0.0153   False  \n",
              "118  0.0151  0.0562    True  \n",
              "119  0.0204  0.0615    True  \n",
              "\n",
              "[120 rows x 7 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-b6fea6f9-ce44-48d7-bd6a-359c32dfd134\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>group1</th>\n",
              "      <th>group2</th>\n",
              "      <th>meandiff</th>\n",
              "      <th>p-adj</th>\n",
              "      <th>lower</th>\n",
              "      <th>upper</th>\n",
              "      <th>reject</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Classic VDB + Cohere Rerank</td>\n",
              "      <td>Classic VDB + HyDE</td>\n",
              "      <td>0.0498</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>0.0293</td>\n",
              "      <td>0.0704</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Classic VDB + Cohere Rerank</td>\n",
              "      <td>Classic VDB + HyDE + Cohere Rerank</td>\n",
              "      <td>0.0221</td>\n",
              "      <td>0.0219</td>\n",
              "      <td>0.0016</td>\n",
              "      <td>0.0427</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Classic VDB + Cohere Rerank</td>\n",
              "      <td>Classic VDB + HyDE + LLM Rerank</td>\n",
              "      <td>0.0599</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>0.0393</td>\n",
              "      <td>0.0805</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Classic VDB + Cohere Rerank</td>\n",
              "      <td>Classic VDB + LLM Rerank</td>\n",
              "      <td>0.0364</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>0.0159</td>\n",
              "      <td>0.0570</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Classic VDB + Cohere Rerank</td>\n",
              "      <td>Classic VDB + Maximal Marginal Relevance (MMR)</td>\n",
              "      <td>0.0006</td>\n",
              "      <td>1.0000</td>\n",
              "      <td>-0.0199</td>\n",
              "      <td>0.0212</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>115</th>\n",
              "      <td>Sentence window retrieval + Cohere rerank</td>\n",
              "      <td>Sentence window retrieval + HyDE + Cohere Rerank</td>\n",
              "      <td>-0.0168</td>\n",
              "      <td>0.2521</td>\n",
              "      <td>-0.0374</td>\n",
              "      <td>0.0037</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>116</th>\n",
              "      <td>Sentence window retrieval + Cohere rerank</td>\n",
              "      <td>Sentence window retrieval + LLM Rerank</td>\n",
              "      <td>0.0241</td>\n",
              "      <td>0.0067</td>\n",
              "      <td>0.0036</td>\n",
              "      <td>0.0447</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>117</th>\n",
              "      <td>Sentence window retrieval + HyDE</td>\n",
              "      <td>Sentence window retrieval + HyDE + Cohere Rerank</td>\n",
              "      <td>-0.0053</td>\n",
              "      <td>0.9999</td>\n",
              "      <td>-0.0259</td>\n",
              "      <td>0.0153</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>118</th>\n",
              "      <td>Sentence window retrieval + HyDE</td>\n",
              "      <td>Sentence window retrieval + LLM Rerank</td>\n",
              "      <td>0.0357</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>0.0151</td>\n",
              "      <td>0.0562</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>119</th>\n",
              "      <td>Sentence window retrieval + HyDE + Cohere Rerank</td>\n",
              "      <td>Sentence window retrieval + LLM Rerank</td>\n",
              "      <td>0.0410</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>0.0204</td>\n",
              "      <td>0.0615</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>120 rows × 7 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-b6fea6f9-ce44-48d7-bd6a-359c32dfd134')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-b6fea6f9-ce44-48d7-bd6a-359c32dfd134 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-b6fea6f9-ce44-48d7-bd6a-359c32dfd134');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-4ed1f7c3-a0f1-4ee7-a518-411d16633608\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-4ed1f7c3-a0f1-4ee7-a518-411d16633608')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-4ed1f7c3-a0f1-4ee7-a518-411d16633608 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "tukey_result_df",
              "summary": "{\n  \"name\": \"tukey_result_df\",\n  \"rows\": 120,\n  \"fields\": [\n    {\n      \"column\": \"group1\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 15,\n        \"samples\": [\n          \"Document summary index + Cohere Rerank\",\n          \"Sentence window retrieval\",\n          \"Classic VDB + Cohere Rerank\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"group2\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 15,\n        \"samples\": [\n          \"Document summary index + HyDE + Cohere Rerank\",\n          \"Sentence window retrieval + Cohere rerank\",\n          \"Classic VDB + HyDE\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"meandiff\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.08812477746559899,\n        \"min\": -0.1291,\n        \"max\": 0.2553,\n        \"num_unique_values\": 115,\n        \"samples\": [\n          0.0542,\n          0.0006,\n          0.164\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"p-adj\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.31182539916277696,\n        \"min\": 0.0,\n        \"max\": 1.0,\n        \"num_unique_values\": 25,\n        \"samples\": [\n          0.9312,\n          0.5672,\n          0.0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"lower\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.08812664987032554,\n        \"min\": -0.1496,\n        \"max\": 0.2347,\n        \"num_unique_values\": 115,\n        \"samples\": [\n          0.0336,\n          -0.0199,\n          0.1435\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"upper\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.08812466865861383,\n        \"min\": -0.1085,\n        \"max\": 0.2759,\n        \"num_unique_values\": 115,\n        \"samples\": [\n          0.0748,\n          0.0212,\n          0.1846\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"reject\",\n      \"properties\": {\n        \"dtype\": \"boolean\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          false,\n          true\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Tukey - Naive RAG"
      ],
      "metadata": {
        "id": "nSniOUsE9NDY"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Filter Tukey HSD results to focus on specific group comparisons\n",
        "filtered_results = tukey_result_df[\n",
        "    (tukey_result_df['group1'] == \"Classic VDB + Naive RAG\") |\n",
        "    (tukey_result_df['group2'] == \"Classic VDB + Naive RAG\")\n",
        "]\n",
        "filtered_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 519
        },
        "id": "J3EdItg9UKGd",
        "outputId": "9f928d42-06c6-436b-fc3b-4e62c93577bd"
      },
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                            group1  \\\n",
              "7                      Classic VDB + Cohere Rerank   \n",
              "21                              Classic VDB + HyDE   \n",
              "34              Classic VDB + HyDE + Cohere Rerank   \n",
              "46                 Classic VDB + HyDE + LLM Rerank   \n",
              "57                        Classic VDB + LLM Rerank   \n",
              "67  Classic VDB + Maximal Marginal Relevance (MMR)   \n",
              "76       Classic VDB + Multi Query + Cohere rerank   \n",
              "84          Classic VDB + Multi Query + Reciprocal   \n",
              "92                         Classic VDB + Naive RAG   \n",
              "93                         Classic VDB + Naive RAG   \n",
              "94                         Classic VDB + Naive RAG   \n",
              "95                         Classic VDB + Naive RAG   \n",
              "96                         Classic VDB + Naive RAG   \n",
              "97                         Classic VDB + Naive RAG   \n",
              "98                         Classic VDB + Naive RAG   \n",
              "\n",
              "                                              group2  meandiff   p-adj  \\\n",
              "7                            Classic VDB + Naive RAG   -0.0150  0.4515   \n",
              "21                           Classic VDB + Naive RAG   -0.0648  0.0000   \n",
              "34                           Classic VDB + Naive RAG   -0.0371  0.0000   \n",
              "46                           Classic VDB + Naive RAG   -0.0749  0.0000   \n",
              "57                           Classic VDB + Naive RAG   -0.0514  0.0000   \n",
              "67                           Classic VDB + Naive RAG   -0.0156  0.3787   \n",
              "76                           Classic VDB + Naive RAG    0.0012  1.0000   \n",
              "84                           Classic VDB + Naive RAG    0.0542  0.0000   \n",
              "92            Document summary index + Cohere Rerank    0.0545  0.0000   \n",
              "93     Document summary index + HyDE + Cohere Rerank    0.0654  0.0000   \n",
              "94                         Sentence window retrieval    0.1679  0.0000   \n",
              "95         Sentence window retrieval + Cohere rerank    0.1769  0.0000   \n",
              "96                  Sentence window retrieval + HyDE    0.1654  0.0000   \n",
              "97  Sentence window retrieval + HyDE + Cohere Rerank    0.1601  0.0000   \n",
              "98            Sentence window retrieval + LLM Rerank    0.2011  0.0000   \n",
              "\n",
              "     lower   upper  reject  \n",
              "7  -0.0355  0.0056   False  \n",
              "21 -0.0854 -0.0442    True  \n",
              "34 -0.0576 -0.0165    True  \n",
              "46 -0.0954 -0.0543    True  \n",
              "57 -0.0720 -0.0308    True  \n",
              "67 -0.0361  0.0050   False  \n",
              "76 -0.0193  0.0218   False  \n",
              "84  0.0336  0.0748    True  \n",
              "92  0.0340  0.0751    True  \n",
              "93  0.0449  0.0860    True  \n",
              "94  0.1474  0.1885    True  \n",
              "95  0.1564  0.1975    True  \n",
              "96  0.1449  0.1860    True  \n",
              "97  0.1396  0.1807    True  \n",
              "98  0.1805  0.2217    True  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-9ad22823-814a-4cab-8b69-65b892d1be7d\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>group1</th>\n",
              "      <th>group2</th>\n",
              "      <th>meandiff</th>\n",
              "      <th>p-adj</th>\n",
              "      <th>lower</th>\n",
              "      <th>upper</th>\n",
              "      <th>reject</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>Classic VDB + Cohere Rerank</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>-0.0150</td>\n",
              "      <td>0.4515</td>\n",
              "      <td>-0.0355</td>\n",
              "      <td>0.0056</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>21</th>\n",
              "      <td>Classic VDB + HyDE</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>-0.0648</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>-0.0854</td>\n",
              "      <td>-0.0442</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>34</th>\n",
              "      <td>Classic VDB + HyDE + Cohere Rerank</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>-0.0371</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>-0.0576</td>\n",
              "      <td>-0.0165</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>46</th>\n",
              "      <td>Classic VDB + HyDE + LLM Rerank</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>-0.0749</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>-0.0954</td>\n",
              "      <td>-0.0543</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>57</th>\n",
              "      <td>Classic VDB + LLM Rerank</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>-0.0514</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>-0.0720</td>\n",
              "      <td>-0.0308</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>67</th>\n",
              "      <td>Classic VDB + Maximal Marginal Relevance (MMR)</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>-0.0156</td>\n",
              "      <td>0.3787</td>\n",
              "      <td>-0.0361</td>\n",
              "      <td>0.0050</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>76</th>\n",
              "      <td>Classic VDB + Multi Query + Cohere rerank</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>0.0012</td>\n",
              "      <td>1.0000</td>\n",
              "      <td>-0.0193</td>\n",
              "      <td>0.0218</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>84</th>\n",
              "      <td>Classic VDB + Multi Query + Reciprocal</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>0.0542</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>0.0336</td>\n",
              "      <td>0.0748</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>92</th>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>Document summary index + Cohere Rerank</td>\n",
              "      <td>0.0545</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>0.0340</td>\n",
              "      <td>0.0751</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>93</th>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>Document summary index + HyDE + Cohere Rerank</td>\n",
              "      <td>0.0654</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>0.0449</td>\n",
              "      <td>0.0860</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>94</th>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>Sentence window retrieval</td>\n",
              "      <td>0.1679</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>0.1474</td>\n",
              "      <td>0.1885</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>95</th>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>Sentence window retrieval + Cohere rerank</td>\n",
              "      <td>0.1769</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>0.1564</td>\n",
              "      <td>0.1975</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>96</th>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>Sentence window retrieval + HyDE</td>\n",
              "      <td>0.1654</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>0.1449</td>\n",
              "      <td>0.1860</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>97</th>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>Sentence window retrieval + HyDE + Cohere Rerank</td>\n",
              "      <td>0.1601</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>0.1396</td>\n",
              "      <td>0.1807</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>98</th>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>Sentence window retrieval + LLM Rerank</td>\n",
              "      <td>0.2011</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>0.1805</td>\n",
              "      <td>0.2217</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-9ad22823-814a-4cab-8b69-65b892d1be7d')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-9ad22823-814a-4cab-8b69-65b892d1be7d button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-9ad22823-814a-4cab-8b69-65b892d1be7d');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-31544948-7912-4b60-9e23-d55a0aaf2743\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-31544948-7912-4b60-9e23-d55a0aaf2743')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-31544948-7912-4b60-9e23-d55a0aaf2743 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "filtered_results",
              "summary": "{\n  \"name\": \"filtered_results\",\n  \"rows\": 15,\n  \"fields\": [\n    {\n      \"column\": \"group1\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 9,\n        \"samples\": [\n          \"Classic VDB + Multi Query + Reciprocal\",\n          \"Classic VDB + HyDE\",\n          \"Classic VDB + Maximal Marginal Relevance (MMR)\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"group2\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 8,\n        \"samples\": [\n          \"Document summary index + Cohere Rerank\",\n          \"Sentence window retrieval + HyDE\",\n          \"Classic VDB + Naive RAG\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"meandiff\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.09856502905382289,\n        \"min\": -0.0749,\n        \"max\": 0.2011,\n        \"num_unique_values\": 15,\n        \"samples\": [\n          0.0654,\n          0.1769,\n          -0.015\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"p-adj\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.28334193594783436,\n        \"min\": 0.0,\n        \"max\": 1.0,\n        \"num_unique_values\": 4,\n        \"samples\": [\n          0.0,\n          1.0,\n          0.4515\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"lower\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.09857018529236632,\n        \"min\": -0.0954,\n        \"max\": 0.1805,\n        \"num_unique_values\": 15,\n        \"samples\": [\n          0.0449,\n          0.1564,\n          -0.0355\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"upper\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.0985650290538229,\n        \"min\": -0.0543,\n        \"max\": 0.2217,\n        \"num_unique_values\": 15,\n        \"samples\": [\n          0.086,\n          0.1975,\n          0.0056\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"reject\",\n      \"properties\": {\n        \"dtype\": \"boolean\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          true,\n          false\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "QwtK55U9LxBz"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Tukey - best classic vdb techniques\n"
      ],
      "metadata": {
        "id": "Ye8NLENr9tiE"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Focus on experiments for a detailed comparison\n",
        "experiments_focus_group1 = [\n",
        "    \"Classic VDB + HyDE\",\n",
        "    \"Classic VDB + HyDE + Cohere Rerank\",\n",
        "    \"Classic VDB + HyDE + LLM Rerank\",\n",
        "    \"Classic VDB + LLM Rerank\"\n",
        "]\n",
        "\n",
        "filtered_results = tukey_result_df[\n",
        "    tukey_result_df['group1'].isin(experiments_focus_group1) &\n",
        "    (tukey_result_df['group2'] == \"Classic VDB + Naive RAG\")\n",
        "]\n",
        "filtered_results\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 175
        },
        "id": "YDxUKcL-UebW",
        "outputId": "ed492f54-97aa-42c1-f166-2708ebd94763"
      },
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                group1                   group2  meandiff  \\\n",
              "21                  Classic VDB + HyDE  Classic VDB + Naive RAG   -0.0648   \n",
              "34  Classic VDB + HyDE + Cohere Rerank  Classic VDB + Naive RAG   -0.0371   \n",
              "46     Classic VDB + HyDE + LLM Rerank  Classic VDB + Naive RAG   -0.0749   \n",
              "57            Classic VDB + LLM Rerank  Classic VDB + Naive RAG   -0.0514   \n",
              "\n",
              "    p-adj   lower   upper  reject  \n",
              "21    0.0 -0.0854 -0.0442    True  \n",
              "34    0.0 -0.0576 -0.0165    True  \n",
              "46    0.0 -0.0954 -0.0543    True  \n",
              "57    0.0 -0.0720 -0.0308    True  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-8249b8fb-5fab-4d3d-8aab-5e0d8230a6ca\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>group1</th>\n",
              "      <th>group2</th>\n",
              "      <th>meandiff</th>\n",
              "      <th>p-adj</th>\n",
              "      <th>lower</th>\n",
              "      <th>upper</th>\n",
              "      <th>reject</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>21</th>\n",
              "      <td>Classic VDB + HyDE</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>-0.0648</td>\n",
              "      <td>0.0</td>\n",
              "      <td>-0.0854</td>\n",
              "      <td>-0.0442</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>34</th>\n",
              "      <td>Classic VDB + HyDE + Cohere Rerank</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>-0.0371</td>\n",
              "      <td>0.0</td>\n",
              "      <td>-0.0576</td>\n",
              "      <td>-0.0165</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>46</th>\n",
              "      <td>Classic VDB + HyDE + LLM Rerank</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>-0.0749</td>\n",
              "      <td>0.0</td>\n",
              "      <td>-0.0954</td>\n",
              "      <td>-0.0543</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>57</th>\n",
              "      <td>Classic VDB + LLM Rerank</td>\n",
              "      <td>Classic VDB + Naive RAG</td>\n",
              "      <td>-0.0514</td>\n",
              "      <td>0.0</td>\n",
              "      <td>-0.0720</td>\n",
              "      <td>-0.0308</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-8249b8fb-5fab-4d3d-8aab-5e0d8230a6ca')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-8249b8fb-5fab-4d3d-8aab-5e0d8230a6ca button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-8249b8fb-5fab-4d3d-8aab-5e0d8230a6ca');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-29ca6a22-a885-4cd5-b7d8-d3c04aefcd4f\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-29ca6a22-a885-4cd5-b7d8-d3c04aefcd4f')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-29ca6a22-a885-4cd5-b7d8-d3c04aefcd4f button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "filtered_results",
              "summary": "{\n  \"name\": \"filtered_results\",\n  \"rows\": 4,\n  \"fields\": [\n    {\n      \"column\": \"group1\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 4,\n        \"samples\": [\n          \"Classic VDB + HyDE + Cohere Rerank\",\n          \"Classic VDB + LLM Rerank\",\n          \"Classic VDB + HyDE\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"group2\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Classic VDB + Naive RAG\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"meandiff\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.01641757188705646,\n        \"min\": -0.0749,\n        \"max\": -0.0371,\n        \"num_unique_values\": 4,\n        \"samples\": [\n          -0.0371\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"p-adj\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.0,\n        \"min\": 0.0,\n        \"max\": 0.0,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          0.0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"lower\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.016421936548409875,\n        \"min\": -0.0954,\n        \"max\": -0.0576,\n        \"num_unique_values\": 4,\n        \"samples\": [\n          -0.0576\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"upper\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.016417571887056462,\n        \"min\": -0.0543,\n        \"max\": -0.0165,\n        \"num_unique_values\": 4,\n        \"samples\": [\n          -0.0165\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"reject\",\n      \"properties\": {\n        \"dtype\": \"boolean\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          true\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Focus on experiments for a detailed comparison\n",
        "experiments_focus_group1 = [\n",
        "    \"Classic VDB + HyDE\",\n",
        "    \"Classic VDB + HyDE + Cohere Rerank\",\n",
        "    \"Classic VDB + HyDE + LLM Rerank\",\n",
        "    \"Classic VDB + LLM Rerank\"\n",
        "]\n",
        "\n",
        "filtered_results = tukey_result_df[\n",
        "    tukey_result_df['group1'].isin(experiments_focus_group1) &\n",
        "    tukey_result_df['group2'].isin(experiments_focus_group1)\n",
        "]\n",
        "filtered_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 237
        },
        "id": "LeTpO2C6L2tS",
        "outputId": "a6ca03e7-0a59-44f7-bb41-842615097ec0"
      },
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                group1                              group2  \\\n",
              "15                  Classic VDB + HyDE  Classic VDB + HyDE + Cohere Rerank   \n",
              "16                  Classic VDB + HyDE     Classic VDB + HyDE + LLM Rerank   \n",
              "17                  Classic VDB + HyDE            Classic VDB + LLM Rerank   \n",
              "29  Classic VDB + HyDE + Cohere Rerank     Classic VDB + HyDE + LLM Rerank   \n",
              "30  Classic VDB + HyDE + Cohere Rerank            Classic VDB + LLM Rerank   \n",
              "42     Classic VDB + HyDE + LLM Rerank            Classic VDB + LLM Rerank   \n",
              "\n",
              "    meandiff   p-adj   lower   upper  reject  \n",
              "15   -0.0277  0.0006 -0.0483 -0.0072    True  \n",
              "16    0.0101  0.9418 -0.0105  0.0306   False  \n",
              "17   -0.0134  0.6445 -0.0340  0.0072   False  \n",
              "29    0.0378  0.0000  0.0172  0.0584    True  \n",
              "30    0.0143  0.5282 -0.0062  0.0349   False  \n",
              "42   -0.0235  0.0102 -0.0440 -0.0029    True  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-67742363-0273-4a81-a9ae-88721036d5c8\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>group1</th>\n",
              "      <th>group2</th>\n",
              "      <th>meandiff</th>\n",
              "      <th>p-adj</th>\n",
              "      <th>lower</th>\n",
              "      <th>upper</th>\n",
              "      <th>reject</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>Classic VDB + HyDE</td>\n",
              "      <td>Classic VDB + HyDE + Cohere Rerank</td>\n",
              "      <td>-0.0277</td>\n",
              "      <td>0.0006</td>\n",
              "      <td>-0.0483</td>\n",
              "      <td>-0.0072</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>Classic VDB + HyDE</td>\n",
              "      <td>Classic VDB + HyDE + LLM Rerank</td>\n",
              "      <td>0.0101</td>\n",
              "      <td>0.9418</td>\n",
              "      <td>-0.0105</td>\n",
              "      <td>0.0306</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>Classic VDB + HyDE</td>\n",
              "      <td>Classic VDB + LLM Rerank</td>\n",
              "      <td>-0.0134</td>\n",
              "      <td>0.6445</td>\n",
              "      <td>-0.0340</td>\n",
              "      <td>0.0072</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>29</th>\n",
              "      <td>Classic VDB + HyDE + Cohere Rerank</td>\n",
              "      <td>Classic VDB + HyDE + LLM Rerank</td>\n",
              "      <td>0.0378</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>0.0172</td>\n",
              "      <td>0.0584</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>30</th>\n",
              "      <td>Classic VDB + HyDE + Cohere Rerank</td>\n",
              "      <td>Classic VDB + LLM Rerank</td>\n",
              "      <td>0.0143</td>\n",
              "      <td>0.5282</td>\n",
              "      <td>-0.0062</td>\n",
              "      <td>0.0349</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>42</th>\n",
              "      <td>Classic VDB + HyDE + LLM Rerank</td>\n",
              "      <td>Classic VDB + LLM Rerank</td>\n",
              "      <td>-0.0235</td>\n",
              "      <td>0.0102</td>\n",
              "      <td>-0.0440</td>\n",
              "      <td>-0.0029</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-67742363-0273-4a81-a9ae-88721036d5c8')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-67742363-0273-4a81-a9ae-88721036d5c8 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-67742363-0273-4a81-a9ae-88721036d5c8');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-5b1f6ebc-084b-4070-94a0-05f128f744a7\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-5b1f6ebc-084b-4070-94a0-05f128f744a7')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-5b1f6ebc-084b-4070-94a0-05f128f744a7 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "filtered_results",
              "summary": "{\n  \"name\": \"filtered_results\",\n  \"rows\": 6,\n  \"fields\": [\n    {\n      \"column\": \"group1\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 3,\n        \"samples\": [\n          \"Classic VDB + HyDE\",\n          \"Classic VDB + HyDE + Cohere Rerank\",\n          \"Classic VDB + HyDE + LLM Rerank\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"group2\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 3,\n        \"samples\": [\n          \"Classic VDB + HyDE + Cohere Rerank\",\n          \"Classic VDB + HyDE + LLM Rerank\",\n          \"Classic VDB + LLM Rerank\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"meandiff\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.02543021824522943,\n        \"min\": -0.0277,\n        \"max\": 0.0378,\n        \"num_unique_values\": 6,\n        \"samples\": [\n          -0.0277,\n          0.0101,\n          -0.0235\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"p-adj\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.407099606566583,\n        \"min\": 0.0,\n        \"max\": 0.9418,\n        \"num_unique_values\": 6,\n        \"samples\": [\n          0.0006,\n          0.9418,\n          0.0102\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"lower\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.02542366351780692,\n        \"min\": -0.0483,\n        \"max\": 0.0172,\n        \"num_unique_values\": 6,\n        \"samples\": [\n          -0.0483,\n          -0.0105,\n          -0.044\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"upper\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.025443479845859658,\n        \"min\": -0.0072,\n        \"max\": 0.0584,\n        \"num_unique_values\": 6,\n        \"samples\": [\n          -0.0072,\n          0.0306,\n          -0.0029\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"reject\",\n      \"properties\": {\n        \"dtype\": \"boolean\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          false,\n          true\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 34
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Tukey - best classic vdb vs worst sentence window"
      ],
      "metadata": {
        "id": "oxwOfQi_-tfb"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Filter the original Tukey HSD results DataFrame to include only comparisons between the two experiments of interest\n",
        "filtered_results = tukey_result_df[\n",
        "    ((tukey_result_df['group1'] == \"Classic VDB + HyDE + LLM Rerank\") & (tukey_result_df['group2'] == \"Sentence window retrieval + Cohere rerank\")) |\n",
        "    ((tukey_result_df['group1'] == \"Sentence window retrieval + Cohere rerank\") & (tukey_result_df['group2'] == \"Classic VDB + HyDE + LLM Rerank\"))\n",
        "]\n",
        "\n",
        "# Display the filtered results\n",
        "filtered_results\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 81
        },
        "id": "s3oG3P-8n-Ov",
        "outputId": "0ac1ef3d-2266-4df6-af6d-ff2d80c7f2cc"
      },
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                             group1  \\\n",
              "50  Classic VDB + HyDE + LLM Rerank   \n",
              "\n",
              "                                       group2  meandiff  p-adj   lower  \\\n",
              "50  Sentence window retrieval + Cohere rerank    0.1021    0.0  0.0815   \n",
              "\n",
              "     upper  reject  \n",
              "50  0.1226    True  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-9cd055e0-626f-4146-b84f-2ceb92acc855\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>group1</th>\n",
              "      <th>group2</th>\n",
              "      <th>meandiff</th>\n",
              "      <th>p-adj</th>\n",
              "      <th>lower</th>\n",
              "      <th>upper</th>\n",
              "      <th>reject</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>50</th>\n",
              "      <td>Classic VDB + HyDE + LLM Rerank</td>\n",
              "      <td>Sentence window retrieval + Cohere rerank</td>\n",
              "      <td>0.1021</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0815</td>\n",
              "      <td>0.1226</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-9cd055e0-626f-4146-b84f-2ceb92acc855')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-9cd055e0-626f-4146-b84f-2ceb92acc855 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-9cd055e0-626f-4146-b84f-2ceb92acc855');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "filtered_results",
              "summary": "{\n  \"name\": \"filtered_results\",\n  \"rows\": 1,\n  \"fields\": [\n    {\n      \"column\": \"group1\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Classic VDB + HyDE + LLM Rerank\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"group2\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Sentence window retrieval + Cohere rerank\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"meandiff\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": 0.1021,\n        \"max\": 0.1021,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          0.1021\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"p-adj\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": 0.0,\n        \"max\": 0.0,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          0.0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"lower\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": 0.0815,\n        \"max\": 0.0815,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          0.0815\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"upper\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": 0.1226,\n        \"max\": 0.1226,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          0.1226\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"reject\",\n      \"properties\": {\n        \"dtype\": \"boolean\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          true\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Tukey - Sentence Window"
      ],
      "metadata": {
        "id": "Tc9KwTnv_kK2"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Filter the Tukey HSD results for comparisons involving \"Sentence window retrieval\"\n",
        "filtered_results = tukey_result_df[\n",
        "    (tukey_result_df['group1'] == \"Sentence window retrieval\")\n",
        "]\n",
        "\n",
        "filtered_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 175
        },
        "id": "8mINcBuMsnDg",
        "outputId": "2b9a89ee-0712-411f-90d2-f8aaff24ce4c"
      },
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                        group1  \\\n",
              "110  Sentence window retrieval   \n",
              "111  Sentence window retrieval   \n",
              "112  Sentence window retrieval   \n",
              "113  Sentence window retrieval   \n",
              "\n",
              "                                               group2  meandiff   p-adj  \\\n",
              "110         Sentence window retrieval + Cohere rerank    0.0090  0.9768   \n",
              "111                  Sentence window retrieval + HyDE   -0.0025  1.0000   \n",
              "112  Sentence window retrieval + HyDE + Cohere Rerank   -0.0078  0.9945   \n",
              "113            Sentence window retrieval + LLM Rerank    0.0332  0.0000   \n",
              "\n",
              "      lower   upper  reject  \n",
              "110 -0.0115  0.0296   False  \n",
              "111 -0.0231  0.0181   False  \n",
              "112 -0.0283  0.0128   False  \n",
              "113  0.0126  0.0537    True  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-6731e7cd-38ae-44b1-afe6-fce8fdb2916c\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>group1</th>\n",
              "      <th>group2</th>\n",
              "      <th>meandiff</th>\n",
              "      <th>p-adj</th>\n",
              "      <th>lower</th>\n",
              "      <th>upper</th>\n",
              "      <th>reject</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>110</th>\n",
              "      <td>Sentence window retrieval</td>\n",
              "      <td>Sentence window retrieval + Cohere rerank</td>\n",
              "      <td>0.0090</td>\n",
              "      <td>0.9768</td>\n",
              "      <td>-0.0115</td>\n",
              "      <td>0.0296</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>111</th>\n",
              "      <td>Sentence window retrieval</td>\n",
              "      <td>Sentence window retrieval + HyDE</td>\n",
              "      <td>-0.0025</td>\n",
              "      <td>1.0000</td>\n",
              "      <td>-0.0231</td>\n",
              "      <td>0.0181</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>112</th>\n",
              "      <td>Sentence window retrieval</td>\n",
              "      <td>Sentence window retrieval + HyDE + Cohere Rerank</td>\n",
              "      <td>-0.0078</td>\n",
              "      <td>0.9945</td>\n",
              "      <td>-0.0283</td>\n",
              "      <td>0.0128</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>113</th>\n",
              "      <td>Sentence window retrieval</td>\n",
              "      <td>Sentence window retrieval + LLM Rerank</td>\n",
              "      <td>0.0332</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>0.0126</td>\n",
              "      <td>0.0537</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-6731e7cd-38ae-44b1-afe6-fce8fdb2916c')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-6731e7cd-38ae-44b1-afe6-fce8fdb2916c button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-6731e7cd-38ae-44b1-afe6-fce8fdb2916c');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-b38d6558-bd57-4957-8e1d-95045791c2b1\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-b38d6558-bd57-4957-8e1d-95045791c2b1')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-b38d6558-bd57-4957-8e1d-95045791c2b1 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "filtered_results",
              "summary": "{\n  \"name\": \"filtered_results\",\n  \"rows\": 4,\n  \"fields\": [\n    {\n      \"column\": \"group1\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Sentence window retrieval\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"group2\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 4,\n        \"samples\": [\n          \"Sentence window retrieval + HyDE\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"meandiff\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.018220203987149356,\n        \"min\": -0.0078,\n        \"max\": 0.0332,\n        \"num_unique_values\": 4,\n        \"samples\": [\n          -0.0025\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"p-adj\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.4953155786970565,\n        \"min\": 0.0,\n        \"max\": 1.0,\n        \"num_unique_values\": 4,\n        \"samples\": [\n          1.0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"lower\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.018193290887943647,\n        \"min\": -0.0283,\n        \"max\": 0.0126,\n        \"num_unique_values\": 4,\n        \"samples\": [\n          -0.0231\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"upper\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.018174065771496113,\n        \"min\": 0.0128,\n        \"max\": 0.0537,\n        \"num_unique_values\": 4,\n        \"samples\": [\n          0.0181\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"reject\",\n      \"properties\": {\n        \"dtype\": \"boolean\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          true\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Tukey - Doc summary vs Classic VDB vs Sentence window"
      ],
      "metadata": {
        "id": "3ZtriZtk-frH"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Filter for broader comparisons across selected experiments\n",
        "filtered_results = tukey_result_df[\n",
        "    tukey_result_df['group1'].isin([\n",
        "        \"Sentence window retrieval\",\n",
        "        \"Classic VDB + Naive RAG\",\n",
        "        \"Document summary index + Cohere Rerank\"\n",
        "    ]) & tukey_result_df['group2'].isin([\n",
        "        \"Sentence window retrieval\",\n",
        "        \"Classic VDB + Naive RAG\",\n",
        "        \"Document summary index + Cohere Rerank\"\n",
        "    ])\n",
        "]\n",
        "print(filtered_results)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rdIEqiojubR6",
        "outputId": "3f004634-f855-4db9-c224-af056cc7aebc"
      },
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "                                     group1  \\\n",
            "92                  Classic VDB + Naive RAG   \n",
            "94                  Classic VDB + Naive RAG   \n",
            "100  Document summary index + Cohere Rerank   \n",
            "\n",
            "                                     group2  meandiff  p-adj   lower   upper  \\\n",
            "92   Document summary index + Cohere Rerank    0.0545    0.0  0.0340  0.0751   \n",
            "94                Sentence window retrieval    0.1679    0.0  0.1474  0.1885   \n",
            "100               Sentence window retrieval    0.1134    0.0  0.0928  0.1340   \n",
            "\n",
            "     reject  \n",
            "92     True  \n",
            "94     True  \n",
            "100    True  \n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Tukey - Doc summary techniques comparison"
      ],
      "metadata": {
        "id": "Z2Yu2e3r-WRY"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Specific pairwise comparison\n",
        "filtered_results = tukey_result_df[\n",
        "    (tukey_result_df['group1'] == \"Document summary index + Cohere Rerank\") &\n",
        "    (tukey_result_df['group2'] == \"Document summary index + HyDE + Cohere Rerank\")\n",
        "]\n",
        "filtered_results\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 81
        },
        "id": "DAyJm9EK1odO",
        "outputId": "726e7cad-864e-4dcd-e0ea-f3234ab27d17"
      },
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                    group1  \\\n",
              "99  Document summary index + Cohere Rerank   \n",
              "\n",
              "                                           group2  meandiff   p-adj   lower  \\\n",
              "99  Document summary index + HyDE + Cohere Rerank    0.0109  0.8935 -0.0097   \n",
              "\n",
              "     upper  reject  \n",
              "99  0.0315   False  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-335ddc60-fdbc-42d4-b9b2-7a18375716dd\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>group1</th>\n",
              "      <th>group2</th>\n",
              "      <th>meandiff</th>\n",
              "      <th>p-adj</th>\n",
              "      <th>lower</th>\n",
              "      <th>upper</th>\n",
              "      <th>reject</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>99</th>\n",
              "      <td>Document summary index + Cohere Rerank</td>\n",
              "      <td>Document summary index + HyDE + Cohere Rerank</td>\n",
              "      <td>0.0109</td>\n",
              "      <td>0.8935</td>\n",
              "      <td>-0.0097</td>\n",
              "      <td>0.0315</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-335ddc60-fdbc-42d4-b9b2-7a18375716dd')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-335ddc60-fdbc-42d4-b9b2-7a18375716dd button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-335ddc60-fdbc-42d4-b9b2-7a18375716dd');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "filtered_results",
              "summary": "{\n  \"name\": \"filtered_results\",\n  \"rows\": 1,\n  \"fields\": [\n    {\n      \"column\": \"group1\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Document summary index + Cohere Rerank\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"group2\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Document summary index + HyDE + Cohere Rerank\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"meandiff\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": 0.0109,\n        \"max\": 0.0109,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          0.0109\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"p-adj\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": 0.8935,\n        \"max\": 0.8935,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          0.8935\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"lower\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": -0.0097,\n        \"max\": -0.0097,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          -0.0097\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"upper\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": null,\n        \"min\": 0.0315,\n        \"max\": 0.0315,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          0.0315\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"reject\",\n      \"properties\": {\n        \"dtype\": \"boolean\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          false\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 33
        }
      ]
    }
  ]
}