calcul-metrique-MDS.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "source": [
        "import requests\n",
        "! pip install nested-lookup\n",
        "from nested_lookup import nested_lookup \n",
        "\n",
        "! pip install pandas\n",
        "import pandas as pd\n",
        "! pip install seaborn\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "import csv"
      ],
      "execution_count": null,
      "outputs": [],
      "metadata": {
        "datalore": {
          "sheet_delimiter": false
        },
        "id": "MS-YmkLZR0Hw"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "La fonction *find_references* ci-dessous permet d'interroger l'interface de programmation HAL (https://api.archives-ouvertes.fr/docs) pour récupérer l'ensemble des notices présentes sur l'archive ouverte correspondantes à un champ parametré (*field*, *field_value*)."
      ],
      "metadata": {
        "datalore": {
          "sheet_delimiter": false
        },
        "id": "UoiOKAv_R0Hz"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def find_notices(field_value, field, increment=0):\n",
        "    notices = []\n",
        "    \n",
        "    # Paramétrage des champs à récupérer, ici nous récupérons ceux qui importent pour le calcul de la métrique MDS\n",
        "    # https://api.archives-ouvertes.fr/docs/search/?schema=fields#fields\n",
        "    flags = \"docid,halId_s,doiId_s,\" \\\n",
        "        \"fileMain_s,title_s,*_abstract_s,*_keyword_s,fulltext_t,domain_s,primaryDomain_s,docType_s,\" \\\n",
        "        \"conferenceEndDate_tdate,conferenceStartDate_tdate,defenseDate_tdate,ePublicationDate_tdate,\" \\\n",
        "        \"producedDate_tdate,publicationDate_tdate,releasedDate_tdate,writingDate_tdate,\" \\\n",
        "        \"submittedDateY_i,modifiedDateY_i\"\n",
        "    \n",
        "    # Interrogation de l'API HAL\n",
        "    req = requests.get(\n",
        "        \"http://api.archives-ouvertes.fr/search/?q=\" + field + \":\" + str(field_value) + \"&fl=\" + flags + \"&start=\" + str(\n",
        "            increment))\n",
        "\n",
        "    if req.status_code == 200:\n",
        "        data = req.json()\n",
        "        if \"response\" in data.keys():\n",
        "            data = data[\"response\"]\n",
        "            count = data[\"numFound\"]\n",
        "\n",
        "            for notice in data[\"docs\"]:\n",
        "                notices.append(notice)\n",
        "            if (count > 30) and (increment < count):\n",
        "                increment += 30\n",
        "                # Boucle pour obtenir l'intégralité des documents liés à la requête d'origine\n",
        "                # La réponse de l'API se limitant à 30 notices pour chaque appel\n",
        "                tmp_notices = find_notices(field_value, field, increment=increment)\n",
        "                for tmp_notice in tmp_notices:\n",
        "                    notices.append(tmp_notice)\n",
        "                return notices\n",
        "            else:\n",
        "                return notices\n",
        "        else:\n",
        "            print(\"Warning : wrong response from HAL API endpoint (\" + field_value + \")\")\n",
        "            return notices\n",
        "    else:\n",
        "        print(\"Warning : can not reach HAL API endpoint\")\n",
        "        return notices"
      ],
      "execution_count": 3,
      "outputs": [],
      "metadata": {
        "datalore": {
          "sheet_delimiter": false
        },
        "id": "RjlbTkHyR0H0"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "La fonction *calculate_mds* ci-dessous permet de calculer la métrique MDS pour une notice donnée (dictionnaire)."
      ],
      "metadata": {
        "datalore": {
          "sheet_delimiter": false
        },
        "id": "03vg94x-R0H2"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def calculate_mds(doc):\n",
        "    score = 0\n",
        "\n",
        "    abstract_penalty = False\n",
        "\n",
        "    # Vérification de la présence d'un titre\n",
        "    if \"title_s\" in doc:\n",
        "        has_title = True\n",
        "    else:\n",
        "        has_title = False\n",
        "\n",
        "    # Vérification de la présence d'un DOI\n",
        "    if \"doiId_s\" in doc:\n",
        "        has_doi = True\n",
        "        if doc[\"doiId_s\"] == \"\":\n",
        "            has_doi = False\n",
        "    else:\n",
        "        has_doi = False\n",
        "\n",
        "    # Vérification d'une date de publication (ou d'une date de conférence, etc.)\n",
        "    if \"publicationDate_tdate\" in doc or \"conferenceStartDate_tdate\" in doc or \"conferenceEndDate_tdate\" in doc \\\n",
        "            or \"defenseDate_tdate\" in doc or \"ePublicationDate_tdate\" in doc or \"producedDate_tdate\" in doc \\\n",
        "            or \"releasedDate_tdate\" in doc or \"writingDate_tdate\":\n",
        "        has_publication_date = True\n",
        "    else:\n",
        "        has_publication_date = False\n",
        "\n",
        "    # Vérification de la présence de mots-clés quelque soit la langue (préfixe _keyword_s)\n",
        "    keywords = nested_lookup(\n",
        "        key=\"_keyword_s\",\n",
        "        document=doc,\n",
        "        wild=True,\n",
        "        with_keys=True,\n",
        "    )\n",
        "\n",
        "    if len(keywords) > 0:\n",
        "        has_kw = True\n",
        "    else:\n",
        "        has_kw = False\n",
        "\n",
        "    # Vérification de la présence de résumé quelque soit la langue (préfixe _abstract_s)\n",
        "    abstracts = nested_lookup(\n",
        "        key=\"_abstract_s\",\n",
        "        document=doc,\n",
        "        wild=True,\n",
        "        with_keys=True,\n",
        "    )\n",
        "\n",
        "    # Vérification de la présence d'un résumé\n",
        "    sub_abstract_penalty = 0\n",
        "    for abstract in abstracts:\n",
        "        # Filtrage si le résumé est égal à \"no abstract\"\n",
        "        if len(doc[abstract][0].split()) < 3:\n",
        "            sub_abstract_penalty += 1\n",
        "\n",
        "    if sub_abstract_penalty == len(abstracts) and (sub_abstract_penalty != 0):\n",
        "        abstract_penalty = True\n",
        "\n",
        "    if len(abstracts) > 0 and not abstract_penalty:\n",
        "        has_abstract = True\n",
        "    else:\n",
        "        has_abstract = False\n",
        "\n",
        "    # Vérification de la présence d'un fichier (texte complet)\n",
        "    if 'fileMain_s' in doc:\n",
        "        has_attached_file = True\n",
        "    else:\n",
        "        has_attached_file = False\n",
        "\n",
        "    # Vérification de la présence d'un domaine\n",
        "    if 'domain_s' in doc:\n",
        "        has_domain = True\n",
        "    else:\n",
        "        has_domain = False\n",
        "\n",
        "    if has_title:\n",
        "        score += 1 * 0.1\n",
        "    if has_publication_date:\n",
        "        score += 1 * 0.1\n",
        "    if has_kw:\n",
        "        score += 1 * 1\n",
        "    if has_abstract:\n",
        "        score += 1 * 0.8\n",
        "    if has_attached_file:\n",
        "        score += 1 * 0.4\n",
        "    if has_doi:\n",
        "        score += 1 * 0.6\n",
        "    if has_domain:\n",
        "        score += 1 * 0.1\n",
        "\n",
        "    return score / (0.1 + 0.1 + 1 + 0.8 + 0.4 + 0.6 + 0.1)"
      ],
      "execution_count": 4,
      "outputs": [],
      "metadata": {
        "datalore": {
          "sheet_delimiter": false
        },
        "id": "YsUKqh3_R0H2"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Le bloc de code suivant permet de lire le fichier *halId_s.csv* qui contient les identifiants HAL des chercheurs de l'Université de Toulon (paramétrés à la date de Septembre 2021). Le traitement consiste ensuite à récupérer les documents liés à chaque identifiant HAL."
      ],
      "metadata": {
        "datalore": {
          "sheet_delimiter": false
        },
        "id": "1FdmhO0tR0H3"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# [Optionnel] Nettoyage du fichier CSV où les notices sont écrites\n",
        "df = pd.read_csv('notices_mds.csv',header=None)\n",
        "df.head(1).to_csv('notices_mds.csv',index=False, header=False)"
      ],
      "execution_count": 5,
      "outputs": [],
      "metadata": {
        "datalore": {
          "sheet_delimiter": false
        },
        "id": "HlnC_LGWR0H4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "halId_s = pd.read_csv(\"halId_s.csv\")\n",
        "\n",
        "# Boucle dans les identifiants HAL des chercheurs\n",
        "for index, row in halId_s.iterrows():\n",
        "    researcher_notices = find_notices(row[\"halId_s\"], \"authIdHal_s\")\n",
        "\n",
        "    # Sauvegarde des éléments pertinents pour notre analyse\n",
        "    for notice in researcher_notices:\n",
        "        # if \"modifiedDateY_i\" in notice:\n",
        "            # notice[\"submittedDateY_i\"] = notice[\"modifiedDateY_i\"]\n",
        "        if 'doiId_s' not in notice:\n",
        "            notice[\"doiId_s\"] = \"\"\n",
        "            \n",
        "        with open(\"notices_mds.csv\", \"a\", encoding=\"utf-8\") as f:\n",
        "            writer = csv.writer(f, lineterminator=\"\\n\")\n",
        "            writer.writerow([notice[\"primaryDomain_s\"], notice[\"doiId_s\"], notice[\"submittedDateY_i\"], notice[\"docType_s\"], round(calculate_mds(notice), 5)])\n",
        "\n",
        "print(\"Récupération des notices terminée\")"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Récupération des notices terminée\n"
          ]
        }
      ],
      "metadata": {
        "datalore": {
          "sheet_delimiter": false
        },
        "id": "r7Ox31LtR0H4",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a9bc9633-42b5-415c-886f-61f13270d74c"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Le bloc de code suivant effectue des opérations sur le jeu de données pour montrer l'évolution des dépôts et de la métrique MDS selon l'année."
      ],
      "metadata": {
        "datalore": {
          "sheet_delimiter": false
        },
        "id": "S4_yNPFVR0H5"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "notices = pd.read_csv(\"notices_mds.csv\")\n",
        "\n",
        "print(\"Moyenne de la métrique MDS des dépôts du jeu de données : \" + str(round(notices[['mds']].mean()['mds'], 3)))\n",
        "print(\"Nombre de dépôts dans le jeu de données : \" + str(len(notices.index)))\n",
        "\n",
        "# Pour des raisons de lisibilité, suppression des résultats précédents l'année 2010 et ceux de l'année 2022 (en cours)\n",
        "notices.drop(notices.index[notices['submittedDateY_i'] < 2010], inplace=True)\n",
        "notices.drop(notices.index[notices['submittedDateY_i'] > 2021], inplace=True)\n",
        "notices = notices.sort_values(by=['submittedDateY_i'])\n",
        "\n",
        "# Calcul des dépôts par année\n",
        "notices_per_year = notices.groupby('submittedDateY_i').size()\n",
        "\n",
        "# Calcul de la moyenne de la métrique MDS par année\n",
        "notices_per_year_mean = notices.groupby('submittedDateY_i').mean()\n",
        "\n",
        "sns.set()\n",
        "\n",
        "notices_per_year.plot(kind='bar', title=\"Evolution des dépôts sur HAL par année\", legend=None)\n",
        "plt.xlabel('Année')\n",
        "plt.ylabel('Dépôts sur HAL')\n",
        "plt.show()\n",
        "\n",
        "notices_per_year_mean.plot(title=\"Evolution du MDS selon l'année de dépôt de la notice\", legend=None)\n",
        "plt.xlabel('Année')\n",
        "plt.ylabel('MDS')\n",
        "plt.show()"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Moyenne de la métrique MDS des dépôts du jeu de données : 0.426\n",
            "Nombre de dépôts dans le jeu de données : 6004\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ],
      "metadata": {
        "datalore": {
          "sheet_delimiter": false
        },
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 638
        },
        "id": "32dqr-1ER0H6",
        "outputId": "5d7ac7ff-b4d4-4f6e-bf1b-9b7e10b97dd0"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Le bloc de code suivant effectue des opérations sur le jeu de données pour montrer l'évolution des dépôts et de la métrique MDS selon la catégorie scientifique (STM ou SHS) à laquelle la notice appartient."
      ],
      "metadata": {
        "datalore": {
          "sheet_delimiter": false
        },
        "id": "4DlqYhbaR0H6"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Informations sur une abbréviation d'un domaine : https://data.archives-ouvertes.fr/subject/shs\n",
        "shs_domains = [\"shs\", \"qfin\", \"scco\"]\n",
        "\n",
        "# Tri des notices\n",
        "notices_shs = notices[notices[\"primaryDomain_s\"].str.contains(\"|\".join(shs_domains)).fillna(False)]\n",
        "notices_stm = notices[~notices[\"primaryDomain_s\"].str.contains(\"|\".join(shs_domains), na=False).fillna(False)]\n",
        "\n",
        "# Calcul des dépôts par année pour les SHS puis STM\n",
        "notices_shs_per_year = notices_shs.groupby(\"submittedDateY_i\").size()\n",
        "notices_stm_per_year = notices_stm.groupby(\"submittedDateY_i\").size()\n",
        "\n",
        "# Calcul de la moyenne de la métrique MDS par année pour les SHS puis STM\n",
        "notices_shs_per_year_mean = notices_shs.groupby(\"submittedDateY_i\").mean()\n",
        "notices_stm_per_year_mean = notices_stm.groupby(\"submittedDateY_i\").mean()\n",
        "\n",
        "# Création du graphique montrant l'évolution des dépôts HAL\n",
        "notices_shs_per_year.plot(kind=\"bar\", title=\"Évolution des dépôts sur HAL par année\\n selon la catégorie scientifique\", label=\"SHS\", color=\"violet\")\n",
        "notices_stm_per_year.plot(kind=\"bar\", title=\"Évolution des dépôts sur HAL par année\\n selon la catégorie scientifique\", label=\"STM\", color=\"olive\")\n",
        "\n",
        "plt.legend(title=\"Dépôts\", ncol=2)\n",
        "plt.xlabel(\"Année\")\n",
        "plt.ylabel(\"Dépôts sur HAL\")\n",
        "plt.show()\n",
        "\n",
        "# Création du graphique montrant l'évolution de la métrique MDS\n",
        "ax = notices_shs_per_year_mean.plot(kind=\"line\", title=\"Évolution du MDS\\n selon l'année de dépôt de la notice\\n et selon la catégorie scientifique\", color=\"violet\")\n",
        "notices_stm_per_year_mean.plot(ax=ax, kind=\"line\", title=\"Évolution du MDS\\n selon l'année de dépôt de la notice\\n et selon la catégorie scientifique\", color=\"olive\")\n",
        "\n",
        "ax.legend([\"SHS\", \"STM\"], title=\"Moyenne annuelle MDS\", ncol=2)\n",
        "\n",
        "ax.set_ylim([0,1])\n",
        "plt.xlabel(\"Année\")\n",
        "plt.ylabel(\"MDS\")\n",
        "plt.show()"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYkAAAE6CAYAAADjg65mAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdd3QV1frw8e/MSYckQEhCSOhICD0QeugJBJIQBCmCqCAo1y42rmLBKyBeFV4RBRv8vAgKSieAoFSRKr2TUEJ6ISG9zOz3j4RjAuk9YX/Wci3PmZl9nj0TznNmz8yzFSGEQJIkSZLyoVZ1AJIkSVL1JZOEJEmSVCCZJCRJkqQCySQhSZIkFUgmCanaS09P59tvvyUjI6OqQ5GkB45MElK195///AcnJyfMzMyqOhRJeuDIJCFVex9++CG+vr7l1p6rqys3btwo1bbHjh1j2LBh5RZLYQYPHszBgwcr5bMkqSAmVR2AJBVk8ODBxMTEYDAYjO89/PDDvPvuu5UWg6urK7/99hvNmjUDwMPDgx07dlTa55eWq6srDRo0YP/+/ZiYZP8zz8zMpH///sTFxXHp0iUAJk+ezMmTJzExMUFRFJo3b46Pjw9PPvmk8cztzp07zJ8/n3379pGSkoKDgwNjxozh6aefrrL+SZVHJgmpWlu6dCl9+vSp6jBqJBsbG/bt28fgwYMB2LdvHzY2NsTFxeVZ791332Xs2LGkpKRw5swZ5s2bx59//smKFStQFIX58+eTkpJCYGAg1tbWXLt2jStXrlRFl6QqIIebpBonIyMDDw8PLl++bHwvLi6OTp06ERsbC8CaNWvw9vamR48ezJgxg8jIyHzbmjx5MmvXrjW+XrduHY8++igAkyZNAiAgIAB3d3cCAwM5fPgw/fv3N64fFBTE5MmT8fDwwNfXl99//924bNasWcyZM4enn34ad3d3xo4dy82bNwvs14YNGxg0aBA9e/bkq6++yrNs1qxZLFy40Pj63jjyExAQwIYNG4yvN27cyKhRowpc38rKyvjZJ0+eZM+ePQCcOXMGf39/bG1tUVWVVq1a4ePjU+hnS7WHTBJSjWNmZoa3tzdbt241vrdt2za6d++OnZ0df/31F59++imLFi3iwIEDODs7M3PmzBJ/zo8//ghkf7meOHGCESNG5FmemZnJjBkz6Nu3LwcPHmT27Nm89tprBAcHG9cJDAzk+eef5+jRozRt2jTPF31uV69eZc6cOXz88cfs37+f+Ph4IiIiShxzbl5eXhw7dow7d+6QkJDAsWPHGDJkSJHbNW7cmA4dOnDs2DEAOnfuzMKFC/n111+5fv16mWKSah6ZJKRq7bnnnsPDw8P435o1awDw9/fPkyQ2b96Mv7+/8f/HjBlD+/btMTMzY+bMmZw8eZJbt26Va2ynTp0iJSWFp59+GjMzM3r37s2gQYPyxOXl5UWnTp0wMTFh5MiRXLhwId+2tm/fzsCBA+nevTtmZma89NJLqGrZ/nmam5szaNAgAgMDCQwMZPDgwZibmxdrWwcHBxISEgB455138Pf358cff8TX1xdvb2/27t1bptikmkNek5CqtSVLluR7TaJnz56kpaVx6tQp7OzsuHjxIl5eXgBERUXRvn1747p16tShXr16REZG4uLiUm6xRUVF0ahRozxf5o0bN84ztNWwYUPj/1tYWJCSklJoW3dZWVlRr169Msc4atQoPv30UwBee+21Ym8XGRmJu7s7kB33jBkzmDFjBklJSXz99de8/PLL7N69u1xilKo3eSYh1UgGgwEfHx+2bNnC1q1bGThwIHXr1gWyfwWHhoYa101JSSE+Ph5HR8f72rG0tCQ1NdX4OiYmptgxODg4EBERga7rxvfCw8Pz/ZzitnVXamoq8fHxeeJMS0srcZweHh5ER0cTExNDt27dirVNeHg4586dw8PD475ldevW5ZlnniElJaXcz8yk6kkmCanG8vf3Z9u2bWzevBk/Pz/j+35+fqxbt44LFy6QkZHBZ599RqdOnfI9i3Bzc2Pnzp2kpqZy48YNfvnllzzLGzZsSEhISL6f36lTJywsLPj222/JzMzk8OHD/PHHH/dduyiOYcOGsWfPHo4dO0ZGRgaff/55nuTj5ubG3r17iY+PJzo6mv/7v/8rVruKorB06VK++uorFEUpdN3U1FSOHDnCs88+S6dOnRgwYACQfTZ3+vRpMjIySE9P54cffsDGxoYWLVqUuJ9SzSOHm6RqbcaMGXmek+jTpw9LliwBsi+oWlpaEhUVledOnz59+vDSSy/xwgsvcOfOHdzd3Qu8YPzEE09w5swZ+vTpg6urK/7+/nkeYHv++eeZNWsWaWlpfPDBB9jZ2RmXmZmZsXTpUubMmcOyZctwdHTk448/plWrViXu50MPPcS7777La6+9RmpqKk8++WSe4aeAgAAOHjzI4MGDcXZ2ZsyYMXz//ffFbrswH3zwAfPmzQOgWbNmDBs2jKlTpxqH0RRF4a233iIsLAwTExNcXV1ZtmwZderUKXE/pZpHkZMOSZIkSQWRw02SJElSgWSSkCRJkgokk4QkSZJUIJkkJEmSpALJJCEVqw5Qady6dQtXV1eysrLKve3i0HWdf/3rX/z888/l1mZ5lu8uqGT5nDlzCrwbqyDJycmMGTOGl19+ucxxLV68uEQP3uVWXcqbL126lLfffruqw6gVZJKQKs3kyZM5fPhwpX3eokWL6NWrF+PHj6+0zyyrn3/+GVNTU1555ZUSbTd//nxef/112rVrV2DBwtoqvx85M2bMYO7cuVUUUe0in5OQaq3SFPWraqVNaB9++CEAvXr1Ks9wJEmeSdQmQgjmzZtH79696dq1K/7+/sZy2hkZGSxYsICBAwfSp08f3n333TxlHnIrz/LXBTl9+jTjx4/Hw8MDT09PPvjggzxzWLu6urJ69WqGDh2Kh4cHc+bM4e4jPXd/HS9YsIDu3bszePDgPAXnEhMTeeutt/D09KRfv34sXLgQTdOMy3/55ReGDx9O9+7deeqpp/KU8LhXYeW7dV3n66+/xsvLi549e/LSSy/lKaVxr2+//RZPT088PT3ve7K7sONz95fy0qVL6dmzJ4MHD2bTpk15+vvGG2/Qq1cvBg0axJdffomu6wQFBfHee+9x8uRJ3N3d8y2zARASEsJjjz2Gu7s7U6ZM4fbt23mWnzx5kgkTJuDh4cHIkSOLfTZY1DHO7e7Q5Pr16xk4cOB9+zsjI4O5c+ca99/cuXPJyMggJSWF6dOnExUVhbu7O+7u7kRGRt43ZHbs2DFjHwYMGMC6deuK3O9SDiHVGvv27RMPP/ywSEhIELqui6tXr4rIyEghhBBz584VzzzzjLh9+7ZITEwUzzzzjPjkk0+EEEIcOnRI9OvXTwghREZGhvDy8hJfffWVSE9PFwcPHhRdunQRQUFBQggh3nzzTdGjRw9x6tQpkZmZKWbOnClefvnlfOMJCQkRbdq0EZmZmfctO3PmjDhx4oTIzMwUISEhwsfHRyxfvty4vE2bNuLpp58WCQkJIjQ0VPTs2VPs3btXCCHEr7/+Ktq1ayd+/vlnkZWVJX788UfRt29foeu6EEKIZ599VrzzzjsiOTlZxMTEiDFjxojVq1cLIYTYuXOn8PLyElevXhWZmZliyZIlYvz48fnGf+XKFdGlSxdx5MgRkZ6eLubNmyfc3NzEn3/+KYQQYsWKFWLs2LEiPDxcpKeni3feeUe88sor+ba1d+9e0bt3b3Hp0iWRnJwsZs6cKdq0aSOuX79erOPj5uYm5s2bJ9LT08Xhw4dF586djcfk9ddfFzNmzBCJiYkiJCREDB06VKxZs8a4ryZMmJBvTHeNGzfO2PaRI0dEly5dxKuvviqEECIiIkL06NFD7NmzR2iaJg4cOCB69OghYmNj821r0KBBxv1T1DHO7e7fyttvvy1SU1PFhQsXRPv27cXVq1eFEEIsWrRIjB07VsTExIjY2Fgxfvx4sXDhQuP+ufv3e9fnn39u7MOtW7dEly5dxObNm0VGRoaIi4sT58+fL3K/S9lkkqhFDh48KIYOHSpOnDghNE0zvq/ruujcubO4ceOG8b2///5bDBo0SAiR9x/Z0aNHRZ8+ffJs/8orr4jPP/9cCJGdJN566y3jsj179ohhw4blG09hSeJey5cvF88++6zxdZs2bcTRo0eNr1988UWxbNkyIUT2F5+Xl5dxWUpKimjTpo2IiooS0dHRon379iI1NdW4fPPmzeKxxx4TQgjx1FNPGb9AhRBC0zTRqVMncevWrftiWrx4cZ4EmJycLNq3b2/8EvTx8REHDx40Lo+MjBTt2rXLt7+zZs0S//3vf42vg4ODjUmiOMfHzc1NJCcn59kfX3zxhcjKyhLt27cXV65cMS5bvXq1sb9FJYnQ0ND72p45c6bxC3bZsmXitddey7PN1KlTxbp16/JtL3eSuNe9xzi3u38r4eHhxvfGjBkjtmzZIoQQYsiQIWLPnj3GZfv27cv37/eu3Eli6dKl+X5uUftdyiavSdQivXv3ZtKkSXzwwQeEhoYydOhQ3nzzTdLT00lNTWX06NHGdYUQeQrI3VWe5a8Lc+3aNT766CPOnj1LamoqmqblKe8NYG9vb/x/S0tLkpOT843B0tISyK72mpCQQFZWFp6ensbluq7j5OQEQFhYGPPmzWPBggXG5UIIIiMjcXZ2zvP5RZXvDgsL47nnnsuzr1RVJTY29r5KsFFRUXTo0MH4OvdnxcXFFXl8bGxssLKyMr5u3LgxUVFR3L59m8zMTBo3bpxnWUEz8d0rKioq37bDw8ONfdy+fTu7d+82Ls/KyqJnz55Ftl2cY3yve4/r3b+tqKio+/oYFRVVrD6Gh4fTtGnT+94vzn6X5IXrWufxxx/n8ccfJzY2lpdffplvv/2WF198EQsLC7Zu3VpkGevc5a/vfvmFh4fTvHnzco3z/fffp127dnz66afUrVuXFStWsGPHjjK326hRI8zMzDh06BAmJvf/eTs5OTFjxgxGjhxZZFsODg4EBQUZX99bvrtRo0bMmzevWCW4HRwcjF+8kP3le1f9+vWLPD537twhJSXF+GUeHh7OQw89RP369TE1NSUsLIzWrVsbl91tp6jKr/b29ve1HRYWZtzOycmJgIAA44XxkijPY+zg4EBYWJixWGF4eDgODg5A0X10cnLi9OnT971fnP0uyQvXtcrp06c5deoUmZmZWFpaYmZmhqqqqKrK2LFjmTdvnnEO6MjISPbv339fG+VZ/rowycnJ1KlThzp16hAUFMTq1avLpV0HBwf69u3LRx99RFJSErquc/PmTY4cOQLAhAkT+Prrr7ly5QqQfdF327Zt+bZVVPnuRx99lEWLFhkvfMfFxbFr16582/Lx8WH9+vVcvXqV1NRUvvjiC+Oy4h6fxYsXk5GRwbFjx9izZw8+Pj7GeTUWLlxIUlISoaGhLF++3JgE7ezsiIyMLPCCsbOzMx06dMjTdu6zhpEjR7J7927279+Ppmmkp6dz+PDhYk2tWp7H2NfXl6+++oq4uDji4uJYsmSJcSZCOzs74uPjSUxMzHfbu5V9AwMDycrK4vbt21y4cKFE/y4eZDJJ1CLJycnMnj2bHj16MGjQIOrVq8dTTz0FwOuvv06zZs0YN24cXbt25cknn+TatWv3tXG3/PW+ffvo1auXcd7l0pS/Lsybb77Jli1b6Nq1K++88065JqGPP/6YzMxMRowYQffu3XnxxReJjo4GwNvbm2nTpjFz5ky6du2Kn58f+/bty7ed3OW7+/Xrh42NTZ7hp8cff5zBgwczdepU3N3dGTduXL6/WAEGDBjAE088wRNPPIG3t/d9t6oWdXwaNmyIjY0N/fr147XXXuP99983HpN33nkHS0tLvLy8mDhxIn5+fowZMwbIviW2devWeHp6FjhE9Omnn3Lq1Cl69uzJkiVLGDVqlHGZk5MTX375JcuWLaN3794MGDCA7777rlhDMuV5jJ999lk6dOjAyJEjGTlyJO3bt+fZZ58FoFWrVvj6+uLl5YWHh8d9Q22NGzfmm2++Yfny5fTo0YNRo0Zx8eJFoPj/Lh5kslS4JFVzhw8f5vXXXy8wmUlSRZJnEpIkSVKBZJKQJEmSCiSHmyRJkqQCyTMJSZIkqUAySUgVqraWIc9PaUp8l8SmTZuYOnVqhbV/L19f3wqr2jtt2jTWr19vfL1w4UJ69uxJ3759CQsLw93dPU+9LanqyIfpJCC7cJ+jo2OJS1Q/SNatW8fatWvzvd+/tCW+S+Lu7Z+VZevWreXSzuLFi7lx4waffPKJ8b1vv/3W+P9hYWEsX76c3bt3Y2dnB8CJEyfK5bOlspNJQpLKQUXPWZGVlZXvE+S1QVhYGPXq1TMmCKl6kcNND5CgoCCmTJlCjx49GDZsGIGBgUD2r+DNmzfz3Xff4e7uzowZM+7bVlTzMuS//vorw4cPx93dnSFDhvDTTz8Vuv6aNWuM648YMYJz584BGEt/331/586dxrjzK7tdVN+/+eYbY3nrtWvX5pmNrqAS35B91jJhwgTmzZtHz549Wbx48X0TCBV0PPOzbt06hgwZgru7+32lxgvaF7lnmSusLHphZb737dvHsmXL2LZtG+7u7sYzocmTJ7N27VoOHjzI1KlTjaW+Z82add9Q4r2lzD/44ANjGfD8hjOLG7dUTFVWWlCqVMnJyaJ///7il19+EZmZmeLcuXOiR48exuqhb775pvjss88K3L66lyHfvXu3uHHjhtB1XRw+fFh06tRJnD17Nt9tAwMDhaenpzh16pTQdV1cv37dWAU2MDBQRERECE3TxNatW0Xnzp2N/cyvomphfd+7d6/o06ePuHz5skhJSRGvvvpqnvLgRZX4dnNzEz/88IPIzMwUqampeT6/qON577F3d3c37ufIyEhx+fLlIvdF7oquhZVFL6rMd+6KrHc99thjxr7eW8X13mNbWCnz/CrAFjduqXjkmcQDYs+ePTg7OzNmzBhMTExo164dw4YNY/v27cXa3sTEhOTkZIKDgxFC0KpVKxwcHBBCsGbNGt566y3q1atH3bp1eeaZZ/Idzz516hQpKSk8/fTTmJmZ0bt3bwYNGpRnXS8vLzp16oSJiQkjR47kwoULxYpv4MCBNG3aFEVR6NGjB3379uXYsWP5rvvLL78wbdo0OnXqhKIoNGvWzFiVdfjw4Tg6OqKqKiNGjKBZs2YFltooqu/btm1j9OjRPPTQQ1haWvLCCy8Yt9U0jcDAQF599VXq1q2Li4sLU6ZMyfML38HBgcmTJ2NiYoKFhUWezy7p8VRVlStXrpCWloaDg4OxUF5h+yK3n376iVdeecVYQPH5559nx44deW4ceP7557GwsKBt27a0bdvWWPqiLMLCwjhz5gwvvfQSZmZmxkmmiqs4cUuFq52DnNJ9QkNDOX36dJ7ZyTRNK/aF0Opehnzv3r0sWbKE69evo+s6aWlptGnTJt91CyodDdkz0S1fvtxYtC8lJeW+mdruKqrU9L3lwe+WKweKVeI7d52oe5XkeFpZWbFw4UK+//573n77bbp27cqbb75Jq1atCt0XuRVWFv2ugsp8l0VRpczLEres/Fo8Mkk8IJycnOjevTvLly/Pd3lR5Zah+pYhz8jI4MUXX2TBggUMGTIEU1NTnn32WeN0p/dycnLK91pHaGgos2fPZsWKFbi7u2MwGAgICDAuv3cfFVVq2sHBIc+Xfu4vtqJKfOf3eff2obDjea9+/frRr18/0tLSWLRoEe+88w6rVq0qcF/cq7Cy6Ldu3Sp02+L8bRWkqFLmlpaWea4BaZpGXFxcseKWikcONz0gBg4cyPXr19mwYQOZmZlkZmZy+vRp43wJdnZ2hf5jr85lyDMyMsjIyKBBgwaYmJiwd+9e/vzzzwLXf+SRR/j+++85e/YsQghu3LhBaGgoqampKIpCgwYNgOyL4XdLisP9ZbeL6ruPjw/r1q0jKCiI1NRUvvzyS2NbRZX4LkpRxzO3mJgYdu3aRUpKCmZmZlhZWRmTdEH74l4lKYt+Lzs7O0JDQ0s1mU9RpcxbtGhBeno6e/bsITMzk6+++ipPWfSyxC1lk0niAVG3bl2+++47AgMD6devH56ennzyySfGf1CPPPIIV69excPDw1iCObfqXIa8bt26zJ49m5dffpnu3buzZcuWQsethw8fzowZM3j11Vfp2rUrzz33HAkJCbRu3ZqpU6cyYcIE+vTpw+XLl+natatxu/zKbhfW9wEDBjB58mQef/xxvL296dy5s3E/QOElvovT58KOZ266rrNixQr69etHjx49OHr0KO+//36h++JeJSmLfi8fHx8AevbsycMPP1ysbXIrrJS5tbU17733HrNnz6Z///5YWlqWupy7lD9Zu0mSKklQUBB+fn6cOXOm1j7zUBnyezhPqjjyTEKSKtDOnTvJyMggISGB//73vwwaNEgmCKlGqZQksWDBAgYPHoyrq6vxAax7aZrGnDlz8PLywtvbm7Vr11ZGaJJUoX766Sd69+6Nt7c3BoPBOMwjSTVFpQw3HTt2DGdnZyZNmsTSpUvzvTVxw4YNbN68mW+++Yb4+HhGjRrFqlWrcHFxqejwJEmSpAJUypmEh4dHnnvE8xMYGMjYsWNRVZUGDRrg5eVV7Ae9JEmSpIpRba5JhIeH53mwyMnJiYiIiCqMSJIkSao2SUKSJEmqfqrNbRZOTk6EhYXRqVMn4P4zi+K6fTsZXS/5ZRY7u7rExiaVeLuaojb3T/at5qrN/aspfVNVhfr16xS4vNokCR8fH9auXcvQoUOJj49n165d/PjjjyVuR9dFqZLE3W1rs9rcP9m3mqs296829K1Shps+/PBD+vfvT0REBFOmTMHX1xeA6dOnc+bMGQACAgJwcXFh6NChjBs3jueee44mTZpURniSJElSAWrdE9exsUmlyt729tZERydWQETVQ23un+xbzVWb+1dT+qaqCnZ2dQtcXm2GmyRJyis1NZnLlyNIT0+v6lAqTFSUWqrCfzVB9eqbgpmZBfXr25e4Kq9MEpJUDaWmJpOYeBs7OwdU1bRM5barMxMTlays6vJFWr6qU9+E0ImPjyEpKQFr63ol2lbeAitJ1VBSUjz16tljbm5RaxOEVHkURcXauj6pqSW/20omCUmqhjQtC1NTs6oOQ6pFDAYTdF0r8XYySUhSNSXPIKTyVNq/J5kkJEmSimnu3Pf5+uvsGQb//vsYDz9ctlkVawKZJCSpBnnkEX8GDuxFfHx8nvenTJmIp6cH4eFhVRSZVBJz576Pp6cH+/fvyfP+559/iqenB4GBmwEIDNxM//498Pbuh7d3P8aOHcm8eXO4efNGnu22bNnAxIlj8Pbuj7//UF577UVSUpLLJVaZJCSphnFyasyuXf9USA4KukpaWloVRiSVRpMmTdm+favxdVZWFn/8sQtn57zTI7Rv35GdO/ezffseFi36EjMzc556ajLBwVcBOHHiOMuWfcn7789l5859/PjjWoYMGVpuccokIUk1zLBhI9i+PdD4etu2Lfj4+OZZJykpif/85138/LwYM8aPFSu+Rdd1MjMzGT58MEFBV43r3r4dx5Ahfbl9+zYAf/65nyefnIiPz0BmzJjK1atXjOs+8og/q1b9jyeemMCwYQN4991/G5/juDv8snr1Svz8vAkIGMbWrZuM22ZkZPDFF4sYPdoXf/+h/Pe/8wpMbqGht3jxxRmMGDEEX98hzJkzm8TEfx5MK0sczz//NJs3bzC+DgzczL/+9ZTx9Y0b13n55WcZPnwwjz46mt9/31mMowIxMdG8/fbr+Pl5MXbsSH7+eXWh6/ft25/Tp09x584dAA4f/otWrR6iQQO7fNc3GAw4O7vw2muzcHfvyvfffw3AhQvn6dChI23atAXAxsaW4cP9sLIquB5TScgkIUk1TPv2HUlOTuL69Wtomsbvv//GsGF5x8YXLvyY5OQk1qzZyBdffM327YEEBm7C1NSUIUOGsmPHP0lm584ddOvWg/r163P58kXmz/+A119/i61bfycgYDSzZs0kIyPDuP7u3Tv59NPFrF27iaCgK2zbttm4LC4uluTkJDZs2MasWe/w2WcLjF+CS5cuJiTkBitWrOLnn9cTHR3N999/k28fhRBMnvwkGzZsY+XKX4iKijR+KZY1jsKkpqbyyivP4e3tw+bNvzFnzjw+++wjrl0LLnQ7Xdd5441XaN26DevXb2PRoi/5+edVHD78V4HbmJmZ4ek5gN9//w2A7du34uNTvGsc/fsP4tSpkwC0b9+Bw4cP8d13yzh9+mSeY1UeZJKQpBoo+2xiK0ePHqZZs+Y0bGhvXHY3cTzzzPNYWdXByakxEyZMMp59DB/ux65dO7hbkWfHjkBjktm0aT0BAaNp374DBoOB4cP9MDU15dy5M8b2H3lkAg0b2mNjY0vfvv24cuWfKYkNBhOefHIaJiYm9O7tiaWlFTdv3kAIwaZN63nxxVexsbHFyqoOjz8+hZ07d+TbPxeXJnTv3gszMzPq16/P+PGTOHnyeJ51ShNHUQ4e3E+jRk74+o7ExMSENm3aMmDAYHbv3lXodhcunCc+/jZTpkzH1NQUZ2cXAgIeNiaAgvj4+LJ9+1YSExM5efJv+vcfWGSMAA0b2nPnTgIAnTu7M3fux1y6dJE33ngZX98hLF78GZpW8ttd8yOfuJakGsjHx5fnnptOeHjofUNNCQnxZGVl0ajRP7NBNmrkRExMNJD9y9PCwoITJ45jZ9eQ0NAQPD37AxAREc62bVv49defjdtmZmYatwXyDIeYm1sQExNjfG1ra4uJyT9fKxYWFqSmphAff5u0tDSeeuox4zIhRIFlK+LiYvl//+8TTp06SUpKCkLoWFvb5FmnNHEUJSIinPPnz+LjM9D4nqZp952p5bddbGxMnu10XadTpy6Fbte5cxfi42/zww/f06ePJ+bmFkXGCBATE4WNja3xde/efenduy+6rvP338d4551ZNGnSjFGjxhSrvcLIJCFJNVCjRk44OTXmr78OMmvWu3mW2drWw8TEhIiIcFq0aAlAZGREnrON4cP92LEjkAYN7EFkB6EAACAASURBVBg4cAjm5uYAODg48vjjU3niiacoT7a29TA3N+d//1uDvb2D8f2CSlcsW7YEUPjhh5+wsbFl3749LFz4cbnEYmlpmedaSGxsrPH/HRwc6dKlK4sWfVmiNh0dHXFyasxPP603vlfcshxDhw5nxYpv+fzzpcX+vH379tC58/0JSFVVPDx60K2bB9euBRW7vcLI4SZJqqH+/e93+fzzr7C0tMzzvsFgYPBgb77++ktSUpKJiAjn559/zPNreOjQ4ezbt4ffftuW50xk5MiH2bhxHefOnUUIQWpqKgcPHijz7ZSqquLv/zCff/4Zt2/HARAdHcWhQwfzXT8lJQUrKyvq1KlLdHQUq1f/UKbPz6116zbs3fsHaWlp3LoVwtatG43L+vbtR0jITbZv30pWVhZZWVlcuHCO69evFdqmm1t7rKysWLlyBenpaWiaRlDQVS5cOFdkPGPHTmDhwiV06dK10PU0TSMsLJSFCz/mxInjTJnyNAD79+9h164d3LlzByEE58+f5eTJv2nfvmMx9kbR5JmEJNVQ994qmdvLL7/OokX/Zdy4AMzMzPH3H4Wv70jjckfHRrRp05bQ0BA6d3Y3vt+2bTveeONtFi78mFu3bmJubk7Hjl3o0sU9v48pkX/96wVWrPiWp5+eQkJCPPb29owePRYPj173rTtlynQ+/PA9fHwG4uzchGHDRrBmzaoyxwAwfvwkLl48j7//UFq3bo23tw/Hjh0BwMqqDgsXfsHixQv54ouF6LqgdeuHeOGFVwpt02Aw8PHHi1i8eCFjxwaQkZFBs2bNmDbtX0XGY2Nji4dHjwKXnzt3Bm/vfgghsLWth7t7N7755geaN28BgLW1Db/88g0LF35MRkYmdnYNefTRyQwdOrwEe6Vgcj6JHDWl9ntp1eb+1ca+RUTcoFGjZhVaSXTevDk0bGjP008/WyHtF0d1qpRa3qpj3+7+XeVW1HwScrhJkh5A4eFh7Nu3Gz+/gKoORarm5HCTJD1gvvnmK9asWcVjjz1J48bOVR2OVM3JJCFJD5jp0//F9OlFj5VLEsjhJkmSJKkQMklIkiRJBZJJQpIkSSqQTBKSJElSgWSSkCRJkgok726SJKlM/vhjF//73/cIARkZ6bRp05b335/LI4/48/HHC2nZsrVx3aeemsxzz71E164eRESE89lnC4iMjEAIgampGW+//V6e9Wub/PZVaOgtMjMzycrKJCTkJi1atAKgTRtXpkyZztixIxk4cAgffrjA2M7cue+zbdsWfvjhpwrfXzJJSJJUajExMXz22Ud8991KHB0bIYTgypVLxdr2008/ok+fvjz88Dggu5ZT7sqttU1B++ruZEHh4WFMmzaZFSv+KT8SHh6GjY0twcFXuXPnDjY2NqSkpHDq1Ik8hRIrUu09IpL0AFAu6agXK6b0g95WRbgWPiIdFxeDwWCCrW297HgUxfilV5SoqKg8X3QV/aV38eIqLl5cWSFtt237GG3bTix0ndLuK0WBIUOGsmvXDkaPHsvu3bvo128ge/f+US6xF0UmCUmSSq116za0a9eeMWN8cXfvRqdOXRg2bITxi3D27DcxMzM3rh8S8s/EP5MmPc4HH7xLmzZtadeuA4MGDcHNrX2l96GyFLWvCuPj48ucObMZPXos27Zt4ZVX3pBJQpKkoglXFa2IX/sVSVVV5s//lODgq5w48Tf79+9h1ar/8cMPPwHw4YcL7rsmcdfQocPp27cvhw4d5tSpv3nxxRm88cbbeHv7VEisbdtOLPLXfkUqbF/lnkAoP87OLpiZmfHXXwdIS0ujVavKu24j726q7WIE6laNzCuZVR2JVIu1bNmaMWPGsWjRl9StW5cTJ44XvRHZkxENGeLNzJlv8sQTT7FrV/7TmdYmpd1Xw4f75ZRPL9482OVFJonaSgiU8zqG9RpKiCBpTRJKcPUqWyzVfNHRUZw9e9r4Oioqkvj42zg5NS5y24MHD5Ceng5gnKTHyan2Fhwsy74CGDTIiwkTym+eiOKSw021UaZA3a+jXhboLgp6PxXz/SB2auiDQTwkfxtI5UPTNL77bhkREeGYm1sghM60af8q1gXZEyeOs2TJIkxMTNA0DVdXN6ZNm1EJUVeNsuwrACsrKyZPfrJig8yHnHQoR62ZuOa2wPCbBrdB91ARXRVQFRra1OX2/+IhHPSBKqJt7UkUtebY5VIZkw5VB7W5f9Wxb6WZdEieSdQiyhUdda8OJqD7qQiXfxKBYq6gjTCg7tAx7NHRdBDtak+ikCSpYsgkURtkCdSDOup5gWgEmpcB6ir3r2eqoPuo8JuOYZ+OpoHoKBOFJEkFq7Qkce3aNWbNmkV8fDz16tVjwYIFNG/ePM86sbGx/Pvf/yY8PJysrCx69uzJ7Nmza/VTmGWWIDDs1FBiQO+ioPdQQc0nQdxloqAPU2GXjuHPnETRRSaK6qiWjQRLVay0f0+V9u3w3nvvMXHiRHbs2MHEiRN5991371tn6dKltGrVis2bN7Np0ybOnTvHb7/9Vlkh1jjKNR3DrxrcAc1HRe9lKDxB3GVQ0L1U9NYKhkM6yvHqNW4qgcFgQmZmRlWHIdUimpaFqhpKvF2lJInY2FjOnz+Pn58fAH5+fpw/f564uLg86ymKQnJyMrquk5GRQWZmJo6OjpURYs2iCdSDGoYdOtiC9ogB0byEh9KgoA9W0dsoGI7qqEc0kL9cq426desRHx9NenqaPKOQykwIncTE21haFnyBuiCVMo4THh6Oo6MjBkN2FjMYDDg4OBAeHk6DBg2M6z377LO88MILeHp6kpqayqRJk+jWrVuJPquwq/RFsbe3LvW2lUVP0Elal4R2S2DuYY6ltyWKSTHOHsi/f2KcIGVrChl/Z2BpZoKllyWKUrz2qpOacOxKxpqEBEsiI6PIzJQPQkployhQp04dmjRxRlVL9oOyWg32b9++HVdXV/7v//6P5ORkpk+fzvbt2/HxKf5j+jXlFlghRIm/jJUQHfV3HTTQvVSSW2sk304q1raF9q+HQM1USD+UTmpiBrqnmv1XVUPUxltgs6m0afNQLe1bttp77Kpn32Jjk+97r6hbYCtluMnJyYnIyEg0TQOyHyqJiorCyckpz3orV65k5MiRqKqKtbU1gwcP5vDhw5URYqVJS4tj/fph/P7708XfSBeoRzXUrTpYgTbGgGhdjodOUdD7quidFdRzIvs22lIkWkmSap9KSRJ2dna4ubmxZcsWALZs2YKbm1ueoSYAFxcX9u3bB0BGRgZ//fUXDz30UGWEWCnS0mLZtGkk4eF/cfnyz9y8ubPojVIE6lYd9bhAuCpoDxugXgX8ylcU9F4qelcF9aJA3SMThSRJlXh30/vvv8/KlSsZNmwYK1euZM6cOQBMnz6dM2fOAPDWW29x/Phx/P39GTVqFM2bN2fcuHGVFWKFSk2NZePGkdy+fYnhw3/C1rYVBw68iaYVcgdLmMDwi4YSIdAGquiDDGBagcNAioLew4DWXUW9LHKGtmSikKQHmSzLkaMixw9TU2PYuNGfhIQghg9fTdOmQ7hxYwdbt46ld+8PcXd/Me8GQqCcFKhHdLABbagB7MqWHEraP+WUjuEvHb25gu6tgqH6XqOojmO/5aU29w1qd/9qSt+qxTWJB1lKSjQbN/qRkBDEiBE/07TpEACaNRtGs2Y+HDv2EcnJEf9skCZQt+sYDuuIFgramLIniNIQnVU0TxX1enY8ZNWq3xKSJBWTTBIVKCUlio0bfblz5xq+vmto0mRQnuWenvPRtAwOHXov+43InOGlEIHmqWb/gjerul/wooOK1l9FCRGo23TIlIlCkh40MklUkJSUSDZu9CUx8Qa+vmtxcRl43zq2tq3o3Pl5Ll1aTeSfhzBszLn7K8CA6FA9bkMV7VT0QSpKmMAQqEGGTBSS9CCRSaICJCdHsGGDL4mJIfj6/oqzc/8C1/Xo9Cp1DE7sP/UGmouO9ogBHKs+OeQmXFX0ISpEgGGrBukyUUjSg0ImiXKWnBzOxo2+JCWF4uf3K87OngWvHCOw2GRJX30OUZzkfLNVYFG9EsRdorWKPlSFaDBs1iBNJgpJehDIJFGOkpLC2LBhBMnJ4fj7r6Nx474FrqtczJ5alExoNXIcTk69OXR4Dunp8ZUYccmIFmp2BdnbOYkiVSYKSartZJIoJ0lJoWzcOIKUlEj8/Nbh5NS7wHXVwxqGPTqikYL2iAHFWaVfv/+Snn6bI0fmVWLUJSeaqdlzUiSAYZMGyTJRSFJtJpNEOUhMvMWGDSNISYnGz289Tk69Cl45U6CcFuitFHRfFayyh5caNuxEu3ZTOHv2G2Jjz1dS5KUjmqhoIwyQRPbF9iSZKCSptpJJoowSE0PYuHEEaWmx+Puvx8mpZ6HrKzcFigainXLf3A89e87GzMyaAwferP7loRsraL4GSMtJFHeqebySJJWKTBJlkJh4k40bfUlLu42//wYaNepR5DZKsEBYgnC6/wK1hYUdPXu+Q2joXoKDN1ZEyOWrkYLmZ4CMnEQRLxOFJNU2MkmU0p07N9iwwZf09Hj8/Tfg6OhR9EaZAuWGQLS4/yzirnbtpmBn14E//3ybzMyUco66AjgoaP4G0HKuUcTJRCFJtYlMEqVw5851NmwYQUZGAv7+G3F0LN7ESMotgZIFomXBt7mqqgmenh+TlBTCiROLyivkitVQQRuZM6HUJg1iZaKQpNpCJokSSki4xoYNI8jMTGTkyE04OLgXe1slSCAsQDQu/FkIZ2dPWrcew4kTi7hz50ZZQ64cDXIShSEnUcTIRCFJtYFMEiWQkBCckyCSGTlyM/b2XYq/cVbOUFPzgoeacuvd+z8oisrBg2+XIeJKVk9BCzCAac5zFNEyUUhSTSeTRDHFx19lw4YRaFoqAQFbsLfvXKLtlVsCJRNEq+I9UW1t7ULXrq8SHLyJkJDdpQm5atjknFHcTRRRMlFIUk0mk0QxxMdfYeNGXzQtnZEjt9CwYccSt6EEC4R50UNNuXXp8gI2Ns1zJifKLPFnVpm7icIcDFtkopCkmkwmiSLcvn2FDRt80fVMAgK20rBhh5I3ogmU6zlDTSWYvMfExIK+fT/i9u2LnD37dck/tyrdmygiZaKQpJpIJolCxMVdYuPGEQihERCwFTu7dqVqR7klUDIKv6upIM2bD6dJkyEcPTqflJToUn1+lbHOuUZhkVM9NkImCkmqaWSSKEBc3EU2bvRFCEFAwFYaNHArdVtKsECYgXApeZJQFAVPz4/Jykrh8OE5pY6hytTNOaOQiUKSaiSZJPIRF3eBjRt9URSFUaMCadCgbekbK+VQU2716z9Ep07PcuHC/4iMPF76WCpBYuItzp9fQUzM2X9Ki9xNFFY5iSK85iYKIfSqDkGSKpVJVQdQ3cTGnmfTJj8UxYSAgK3Ur/9QmdpTwgRKOuilGGrKzcPjDS5f/pkDB15n9OhdKEr1y+8REUfYtu1RUlOzh8WsrBrRpMlgmjYdgovLYCxHNsCwScOwVcuu+5RPaZLqJisrjYiIQ9y6tZdbt/YQE3MKb+8VtGo1sqpDk6RKIZNELjExZ9m0yR+DwYyAgC3Uq1e2BAE5D9CZlm6oKTczMxt69ZrDH3/M4NKl1bRtO6nMsZWnK1d+5Y8/ZlCnTmN8fFYSHx9ESMgurl/fxqVLqwAFe/suNHEZTPPrg2i0pTv4WkAJ7vaqDLqeRXT0iZyksJeIiENoWjqKYsDBoRvm5g04d+5bmSSkB4ZMEjkiIk6xaZMfBoNFToJoXfZG9ZyhpmYKmJT9y9DVdQLnzn3HX3+9R4sWfpib25Y9xjISQnD8+CccOfIfnJx64+OzCktLO5yceuPm9hi6rhEdfYKQkN+5efN3TpxdxN/iU0ypi8vG/jTtOASXjl7Y2raosvjj4i4QGpqdFMLCDpCRcQcAO7sOdOgwDWfnATRu3AczMxuOHp3P0aMfkZh4C2trlyqJWZIqk0wSZD9JvW7dYExMrAgI2IKtbctyaVcJEyhpoBfzAboi21OyJyf65ZdBHDv2MX37zi2XdktL0zLYs+dFLl1aRZs24xk06AsMBvM866iqAUdHDxwdPfDweJP09ARCQ/dxM3gXt67+zrXTgXAabGxa0LTpEJo08cLZuR9mZtYVFvedOze4dWsvoaF7uHVrH6mpUQDY2DSndevRODsPwNm5P1ZW9vdt26bNBI4enc/lyz/RrdtrFRajJFUXMkkAKSlR2Nu3o1+/JeX6i1YJFggTEE3Kb0jFwaErbm6Pc+bMV7Rr9wT167cpt7ZLIi0tlu3bJxMWdoDu3d/Cw+NNFKXofpqb29KypT8tW/oj+ugkbrjMzcQ/uGG5m4sXV3H27LeoqimNGvWkSZMhNG06hIYNO5XpGkxqakxOUsi+rnDnznUALC0dcHEZgIvLQJyd+2Nj06zItmxtW+Dk1IdLl1bRteurxeqzJNVkiqj2s9uUTGxsErpe8i7Z21sTHZ1YfoHoAsP/NERjBd3bUH7tAikp0axa1RVHx274+a0v1hdVefYvPv4qW7eOJTExhMGDv6RNm3GlbyxVZJfvSIBM70zCTY9y8+bvhIT8TkzMaQAsLRvi4jLImDSsrBzzNHFv3zIyEgkLO5CTGPYRG3sWyL6u07ixpzEx1K/ftlRf8ufP/8CePc8zevSuYs0hUhbl/ndZzdTm/tWUvqmqgp1d3QKXyzOJCqKEC5TUst/VlB8rK3u6d/83f/45i+vXA2nRwrfcP6MgoaEH2L59EoqiEhCwpfCpWovDMns+CsMWDdOdprj4eOLcuz+9e88hJSWSkJA/uHnzd27d+oMrV9YCYGfXMWdoajBOTr3JyjIjNHQ/t27t4datvURFHUcIDYPBAienXvTs+R4uLgOwt++Cqpb9T75161EcOPA6ly6trvAkIUlVTZ5J5CjvrK/u11AuCbQnDGBa/olC0zJZu9aTrKxUJkw4gomJRaHrl0f/Ll5cxZ49L2Bj0wJf3zXldu0GgLScM4p40IepiKZ5h5eE0ImJOW1MGhERh9D1TExMrABBVlYqiqLi4NAVZ+eBuLgMoFGjnkXulxJJFCjXBOIhhZ0HpnHz5k6eeOJy+X7GPWrKr9HSqs39qyl9k2cSVUHkfJk0VSokQQAYDKZ4ei5g06aRnDz5OR4eb1TI50D2F/SRI3M5fvy/ODsPYNiwH7CwqF++H2LxzxmFul1HHwai2T+JQlFU7O27YG/fha5dZxqHlEJC/sDKygI7uz44OfUp/zu+hIAIUM/oKNcEigD9joKr60SuXFnL9evbaN364fL9TEmqRmSSqAgRoKRUzFBTbi4uA2nZMoC///4UV9eJFXJLZlZWKn/88S+uXl2Hm9vj9O+/EIPBtNw/B8hOFH45iWLH/YkiNzMza5o3H07z5sMr5hdblkC5KlDP6igxZJdV6aQg7oByUeDSbQB16jTm0qXVMklItVr1e2y3FlCDdYSB7DOJCta371yEEPz11+xybzslJZqNG/24enUdvXv/h4EDF1dcgrgr54wCO1B36CjXK7kMRrJAPaJhWKlh2KODBlp/FW2yAb23Ab2bipIFhksqbdqM5+bNnaSkRFVujJJUiWSSKG9CZN/62kQBs4pPEtbWTena9RWuXl1HaOj+cms3Lu4iv/46mNjYswwbthJ395cq73ZP8+wzCuxA/U1HuVYJiSJSoO7SMPyoofwtEI4Kmp+KNs6AaKf+M2zYUEFvrKCe1XFt8yhCaFy5sqbi45OkKiKTRHmLAiW5dGXBS8vd/WWsrZuyf/8b6HpWmdsLCfmDdeu80LQ0AgICq6YExd1E0RDUnRWUKDSBckXHsC4Lk/Va9vSy7RW0Rw3oww0IFxXySYyio4KSBHbxbXBw6MrFi6vLPzZJqiZkkihnapCOUMkuxVFJTEws6dNnHnFx5zh37vsytXXu3HK2bBlD3bpNGDPmDxwdu5VTlKVgrmQXArTPSRTB5ZQoUgXKcR3DjxqG33VIB61vzpBSXwPYFn7sRDMFYZN9MdvV9VFiY88QE3OmfGKTpGqm0pLEtWvXGD9+PMOGDWP8+PFcv3493/UCAwPx9/fHz88Pf39/YmJiKivEsss91GReuU/itmzpj4vLQI4c+ZDU1NgSb6/rGn/++TZ7975EkyaDGT16B9bWTSog0hK6N1EElSFRxAjUPzQM/9MwHNURdgracBVtggHRUS3+8KCqoHdQUSLgoXqPoKqmOUUMJan2qbQk8d577zFx4kR27NjBxIkTeffdd+9b58yZM3zxxRd8//33bNmyhVWrVmFtXXE1fMpdNChJlTvUdFf25EQLyMhI5PDh/5Ro28zMZHbseIxTpxbTsePTjBjxM2ZmNhUUaSmY5SQKR1B36ShXS5AodIESpGPYkIXJL1p2EndTyBpvQPc1ZN89VYprLaKtgjAFq8v1aNbMh8uX19SsecglqZgqJUnExsZy/vx5/Pz8APDz8+P8+fPExcXlWW/FihVMnToVe/vswmrW1taYm5vf1151pQbnDDU1r5p6Pg0auNGx4zOcP7+c6OhTxdomOTmcDRuGc/36Njw9P6Zfv0/K5ankcmemoI0wQCNQf9dRrhSRKNIEygkdwyoNw04dkkHrnTOk1M8A9ct4jMwURFsFJUjQtnn2HBohIbvK1qYkVUOVkiTCw8NxdHTEYMiuYWQwGHBwcCA8PDzPekFBQYSEhDBp0iQefvhhvvzyS2rMA+F3h5qcK3+oKbfu3WdhaWnH/v2vF7nvYmJO88svg4iPv8rw4T/RqdOMSoqylHInij90lMv5JIo4gbo35xbWwzrCRkEbpqI9akB0Vsv12OgdVdChecIQLCzsuHTpp3JrW5Kqi2r1k1HTNC5dusTy5cvJyMhg2rRpNG7cmFGjRhW7jcIeLy+KvX3ph7aywrNIvJNInf6WmNtX5dmPNV5eH7F58zQiIjbTqdM/kxPl7t/ly1tYv34Clpb1mTr1AI0ada6KYEtFPC5I+imJrN1ZWFlbIOwEtnHmpB9JJ+u6BiZg1tEMi+4WGBzLt7hiHvaQ5JqEekmhU4eJHD+xjLp1s7C0LN+n0cvyd1kT1Ob+1Ya+VUqScHJyIjIyEk3TMBgMaJpGVFQUTk5OedZr3LgxPj4+mJmZYWZmxpAhQzh9+nSJkkRV1W5Sj2soCtyxS4fojFK3Ux5cXB7B3n4JO3a8hp3dIMzMrI39E0Jw+vRXHDz4Fg0bdmbEiJ8xGBrViBozeXgJ1O0KyRtTSNuThp6gI+qC3lPNvuZgoZNCCkRXcByuApNLguZiDEe0xRw69H906PBUuTVfU+r/lFZt7l9N6VtRtZsqZbjJzs4ONzc3tmzZAsCWLVtwc3OjQYMGedbz8/PjwIEDCCHIzMzk0KFDtG3btjJCLJvcQ02WVT+/wN3JiVJSIjh+/FPj+7qexf79r/Lnn7No3nwEo0YFUqdOoyqMtAxMFXQfFdFcQbVV0bxVtIkGhLsKFpV4DJxA2IHDtY40aNBO3uUkVTpNyyA6+lSF3ThRaXc3vf/++6xcuZJhw4axcuVK5syZA8D06dM5cyb7HnNfX1/s7OwYMWIEo0aNonXr1jzyyCOVFWLpxYGSUDV3NRWkUaMeuLpO5NSpxcTHXyU9/Q6BgeM4e/ZbunR5CR+flZia1qnqMMvGVEH3MWD9hDWilQpqFex/RUHvpKLGK7g6TiAy8ii3b1+p/DikB0pmZhJXr65n586pLF/ekrVr+xEWdqBCPkuWCs9RllND9Wh2KQftcUO1OJO4KyUlkh9/7Iq9fReysm4THX2BAQMW0q7dk1UdWrmq8tN6TWBYqZFUP5IVoe1xd59Jr1733+JdGlXetwpWm/tX3n1LTY3h2rVArl3bzK1be9C0dCws7GjefAQtW/rTrNmwUpXOkaXCK4ESLBBO1WOoKTcrK0e6d5/FwYNvY25ui5/fOpo0GVTVYdU+BgW9vYr1MUeaNBrM5cs/0bPn7DJNuSpJAImJNwkO3sy1a1sID/8LIXSsrZvSvv1TtGjhh5NTrwq/ZV0mibKKEyi3QW9fvRLEXR07PoOmZdCt23ig/EuJS9lEewXxN7Q1mcBvSdMIDd2Hi8vAqg5LqmGEEMTFXeDatc0EB28hJib7eacGDdrRrdtrtGjhnzPneyWW/am0T6qllGCBoHpdj8jNYDCjW7dXa/VpfbVgqSAeUmh1dThmZrZcurRaJgmpWITQiYg4yrVr2WcMCQnBgEKjRj3o3fs/tGzph61tqyqLTyaJMlKDdXACrKpnkpAqj95RxfSSJa3tR3E56Bf69/8UU9PSP7cj1V6alkFo6D6Cg7dw/fpWUlIiUVVTnJ3706XLizRvPqLa3Hkok0RZxAuUuOwKopJ0d64Jt7jxnM/6P4KCNtK27aSit5MeCJmZSdy8uYvg4M3cuPEbGRkJmJjUoVkzb1q08KNZs6GYm9er6jDvI5NEGShB2XdRiRbyLELKJjopNN7WE1urlly6tFomiQdcamos168HEhy8mVu3dhvvSGrZciQtW/rh4jIQExPLqg6zUKVKEsHBwQQFBeHm5oaLy4N7MVQN1hGOQF2ZJKRsoqkCtgpts8ZzOHQ+iYk3sbZuWtVhSZUgPT2e+Pgg7twJJj4+iKioP7l5cz9C6NSt24T27afSooV/pdyRVJ6KjHT+/Pm0a9eOgIAAADZs2MBbb72FjY0NKSkpLF68mAEDBlR4oNVOgkCJza4sKklGqoLeUcXtwHgOM59Ll37Cw+ONqo5KKidpaXEkJASRkBBMQkKwMSkkJASTlpa3qrW9fXu6dn2Vli39adiwc6XekVSeikwSu3bt4oknnjC+/uyzz3j77beZNGkS69evZ8mSJQ9kklCCc4aaquldTVLVEa4K1keb0ljpy6VLq+nW7fUa+wXxoBFCkJYWmycR5P7/9PT4XGsrWFs3wda2ViRufwAAIABJREFUJS1bjsLWtiW2ti2pV68VNjbNcXJyqBV3FBaZJG7fvk3jxo0BuHz5MvHx8YwdOxaAkSNHMn/+/IqNsJpSg3WEA2At//FL98iZa8Lt9AR+T32BiIgjODn1rOqopBxCCFJTo3MlgLtJ4BoJCcFkZCQY11UUFWvrptjYtKB16zHUq9cKW9tW2Ni0xMamGSYmFlXYk8pRZJKwtrYmJiaGhg0bcuzYMTp06ICZmRkAWVlZNWe+h/J0R6BEg9ZLDjVJ+dM7qLQ+PZK96ptcurRKJokqFhX1NydPfk58fHZCyMz85xe+ohiwtm6KrW1LGjXqjo1Ny5yzglbY2DTDYDCrwsirXpFJYvjw4bzyyit4e3uzfPlypk+fblx26tQpmjSpBvMgVzI51CQVyUbBtIUNra77cfXqOjw9F1T/X50xIrvkZ4Pa9XcdFLSBXbuexsysLvb2XXFy6p1naKhu3aYYDKZVHWa1VWSSePXVV1m2bBkHDx5k3LhxTJgwwbjswoULeV4/KNRgHWEP2NSuf0xS+dI7qbgFT+CStobr1wNp3Xp0VYdUsDiBYYMGpqBNMFTp7IrlRQjB339/xuHDc2jUqCfDh6/G0rJhVYdV4xSZJExNTXn++efzXZb7gvYDI1GgRIHWUw41SUVoBM4N+1E3tjEXL66qvkkiQ2D4LXtGP9JAPaqje1bgjH6VQNMy2Lv3JS5e/JGHHhrLoEFLqv+ZXDVVrJt109LS+PXXXzl+/DgJCQnY2tri4eHB6NGjsbB4sHa8ck0+QCcVk6JAZ1Pa7hrP8ZDPSUmJxMrKsaqjyksI1H06JIDup6JcEyjnBLgKsK+Zf+NpaXFs3/4YYWEH6N7933h4zJJ3l5VBkT+Hk5KSGDt2LF999RWmpqa0a9cOExMTvvzyS8aOHUtSUlJlxFltqME6wg6oJ//opKKJVgquVuMRQuPy5TVVHc59lHMC9apA764inFX07ipYgOGABjXwppT4+Kv8+usQIiKO4OX1Ld27/1smiDIq8kzi66+/pn79+vz000/UqfPPTGbJyck8//zzfP3118ycObNCg6w2kgRKBGjd5VCTVEwGhXod2+J4uCsXz62ic+fnq8+XVqRAPaijN1UQ7gq3bu1F09Jp3svr/7d359FRlPn+x99PVQLZ2BICJCwTFhPDvss6EyESwEAOIANmdFxGdPwN1xm3EfQOiN5Z+OnR68Lcq7/rMNeBQTAOQUNYBmJE1MgmawiyJGyJARIia7aq5/dHh4wBmpDQ3el0vq9zPMfQXd3fL6Hr0/VU1fNgfmqjDmj07V5S6004eXIza9f+DKUMkpLSiIgY1tAl+YRa93affvopv/3tb2sEBEBwcDBPP/00n376qduK8zbVQ03dG88HRzQ83VNxu3Evxd/v48yZ3Q1djkOpxvynBUFgjzEoPLWdtLRprF59D+mHf8a58BMYWTaUNo6jiZycpXzySRJBQe2YNi1DAsKFag2J/Px8oqOjr/tYdHQ0J0+edHlR3so4YqNDkaEmUTeBittum4ZBMw7s+3tDV+M4D7HRhktgjTO5ZJ9h7dr7CQ7uwNChv+P4iQyWFo1g1+V34OuKhq72hrS2ycp6iYyMx4mMHMXUqf+kVauuDV2WT7mpcZMrN89d78+95tDZ3S5pKAC7mww1ibpr1j+MriRw8NsPsayG3fGqHRrjuMYeaWC3tfjnPx+itLSI8eOXMHjws8yc+TURkcPYpOeSsm8cZ77d1aD1OlNZeZn16x9kx45X6dnzIe6+O8Urp9pu7Go9J1FWVsYbb7xx3ce01pSXl7u8KG+kcjUKuYFO1FOY4va293L4zCccz/snUd0nNkgZ6oSNsc3G7qHQPRVZX73IyZObGDPmvwgP7w9Ay5ZR3H33RxzKSWFz5hw+3BBHv9P/hyFDn8ffP7iWd/CMS5cKSU+fyalTOxgx4g/06/erpvOF1cNqDYnExES+++67Gz7eFKgjGt0aaNPQlYjGqvOQuwhc05YD3/y9YULigsbYYENrsH9icOjwSnbufJPevWdds+6FUorbYqfTRY8lK3MeO3e9xeEjq/jxj1/jRz8a5/naf6CoaB+rV/+U0tIiJkxYRteuDRO4TUWtIfHEE094og7vdlmj8jV6oHJc+y5EPRhRzYhpfg+7T/2F0tJiAgJCPffmlsbcYEGl4zxE8fkcMjJ+RYcOdzBypPNJOpvHtuHOQ28Qc2oGnxpPsXr1PfToMZVRoxY2yD0fR4+uZ/36h2jWLIQpU9ZWH/0I96k1JMaOHVvjZ6119WHdlf/fv3+/e6rzEipXozRYcj5C3AqliO6VzM4d/82hHSn0HvGox97a2GI7Lt+ONygLOsealGT8/YNJSHj/xhPYKYU9yqBjynBmRnzG9pi32LbtFY4fz2DYsAX07PkASnnmc7Fnzzts3vwcYWF9mDhxOSEhkR5536au1pCIiYmhtLSUKVOmMHnyZNq1a+eJuryKOqLRrQAPfvETvqntoH6EfdOLAznLPBYS6oiNsUtj91LY3WHjmsc4f/4okyenERwcUfsLhCp0H4X/rmYMnvJbevSYSmbmb/jss19z4MAy4uLeIDQ01m3123YlX3wxlz173iEqaiJ33fU/+PuHuO39RE21fgVITU3lzTffpKSkhHvvvZdHH32U1atXU1FRgWmamGbjnuOlVqUadVI7TljLUJO4Vf6K2yNmUli6nbMnD7j//b7XGJmOCSntEQbbt79KXl46I0b8gcjIETf9MvZgAx0M5ucWrVv2ICkpjTFj/ouzZw+wYsUovv76JSorS11efnn5edLTZ7Jnzzv06/dvjB+/VALCw27qODE6OprnnnuOjIwMHnzwQTIzMxk1ahT79u1zd30NTuU5hprk0lfhKrcNn4HC5EDWUve+UWXVxH3KcR7i6Il/smXL77nttp/Sp89jdXstf4U9wkCdAZXtGGa+/fafkZy8nR49prF9+6ssXz6MEyc+c1n5588f5x//GMfx4xv5yU/eYOTI32MYPv6l1AvVac+Xl5fH1q1b2blzJ7GxsbRs2dJddXkNdVijWwAyw7BwkaD2HegSNIZvT63ALqt02/sYm21UkeOO6u/tPDZseISwsN7Exb1Zr8tFdTeF3VFhbLHhsuNO7MDAtsTHv8ukSavQWvPxx5PYuPExLl8uuqXaCwu3k5JyJxcunCAx8R/06vXQLb2eqL9aQ6KkpISlS5dyzz338Ktf/YqgoCCWLFnC3/72N99fcKhMhpqEe8T0SuaCzid/6ya3vL7KsTFyNPZARXlkKWvX3gfA+PFL8PcPqueLKuzRBlSC8ZVd46HOne9k5swsBg58hoMHP2TZskHk5Cyt18qVhw+nsmrVRPz9g5g6dQOdO99Zv3qFS9R64nr06NF06tSJpKQk+vXrB8DRo0c5evRo9XOGDx/uvgobkMrTKBus7jLUJFwrqv9Emm9rxYGcv9Np5J2u/RJyRmN8bmN3VFiDFJkZT1BUtJe77/7w1qesaK3Q/RTGNxo7VkPEv+r28wtk2LB53HbbPXz22a/JyHicAwc+4Cc/eZ3WrXvU+tKySJB3qjUkwsPDKSsrY8WKFaxYce1Ux0opNm7c6JbiGpo6otEhQHhDVyJ8jZ9/ID0ip3Lg5HJ+nHsO/26tXPPCZVXnIZqDHW+wZ987HDy4gqFDX3DZTXD2QAN10MLcbGFNM8GoGXBhYT2ZMmUd+/YtJivrRZYvH86gQc8yYMBvnF5uK4sEea9aQyIjI8MTdXifMo06rtG9ZahJuEfM4GT2nVzM4a2ruL3bz2/9BbXjSibOgzXZpKAkiy+/fJ6oqAkMGvTsrb/+Ff4Ke6SBuc5G7dXovtd+PpQy6N37F3TtejebNz/Hli3/wcGDKcTFvXnNDK0+uUhQoab8dDm01I1+KVgZR3FCHXUMNclVTcJd2kcOpVVAdw4ULYPiW5+SW+3WGLkae5jBxZbfsW7dz2nR4keMHfuuy29401EKu7PC2GrDRee1Bwd3ICHhf5k4cQUVFRdZuXIcmZm/pqysBICiooO+t0jQeY252uLiRxcx/2phrqpEfWPDGd0oF3KSPaAT6ohGBwNettqk8B1KKWJi7+UkX3BhR+6tvViBxvjaxu6qqOxVwbp1P6e8/ALjxy+leXMXDWX9UNWd2Fg41p2oRVTUeO6992v69ZvN/v3/y9//PpgdO17jvfeGUVZWQlJSGtHRP3V9nZ6mNcanNmgI/mkweoCCCjC/tvFLsTCXWBiZFuqIDeWNIzAkJK6nvGqoqasMNQn3iuk9E4ADhz+o/wI/l6vmZQoBO87giy+f57vvvmbMmEWEhfV0YbVXaeVY0c44qCG/9tr9/UMYOfIP3HNPJiEhHcnKepHg4PY+tUiQ2q0x8h3TsDeLaYY91MS6x4/K+02sOAPdXqGOaMz1NuZfLYyPLdRO23Ek6aVHGR4LidzcXGbMmEFCQgIzZswgLy/P6XOPHDlCv379WLhwoafKq0Ed0ygLbLmqSbhZixZd6NhuNDnWcthX+zfya9hVM7uWOm6Yy8ldxt69/49+/f6NHj2mur7gq99+gIFu4bgTG+vmdnLh4f2ZNi2DiROX84tffOU7iwQVa4wtNnaUQsdc9eUyWKFvN7DHmVgPmFRONtF9FapMY2bZ+K2wMJdaGJssVJ4NFd4TGB7bC86fP5/k5GTWrVtHcnIy8+bNu+7zLMti/vz5xMfHe6q0a6gjGh0EdGiwEkQTEtM7me/JpXD3Vze9o73C2G5jnNTYow1O69189tlviIwczfDhC9xU7VX8HCex1VlQe2++dsMwiYqaQECAG4bCGoKlMTMs8HdMw37DEQhTQaTCHmZiTfej8j4T68cGOlyhDmrMtTbmYgvjEwu1y4azDXuU4ZGQKCoqIjs7u3rticTERLKzsykuLr7mue+++y5xcXFERUV5orRrVWjUMRlqEp7TvXsSfmYwOZc/QB25+Z2BOmajtmvsGMXlqLOsXXsfAQFhjBv3Vwyj1gsXXUZHGdg/UhjbbLjgPd+APcnYYaPOVAVEYB33GyEK3dPATjCxHjSxJhnoPgp1SWN+ZeO33ML8u4XxuYU66vmjDI/8SyooKKB9+/bVkwGapkm7du0oKCggNPRfU6vm5OSwefNm3n//ff785z/X673Cwuo/+Vd4eAvKs8u5WHmRFgOD8Q/3r/dreaPw8BYNXYLbNO7eWtCz1zRy9qQyJvv/0mZ4uxpX+FyvN/t7m3OfnsNoZxKcFMSaD2dw6VIBDz30OR07dvNk8QBYkyzO/fc5AncYhEyr22ewcf/uoPJkJed3nKdZv2YED625cl+9eusAVC2TYZVYVB6upOJgBRXfVsA+DSb4Rfnh38Mf/+7+mGHunc/Kc183alFRUcHvfvc7/vjHP97SzLJFRRew7bonbXh4C06fPo+x00IFQEngZTjt+lktG8qV/nyRL/QWFTWd3bvf52D+J3Tb+1Po4AiJ6/ZmacxVjgWEKsbAZ+vncvjweuLi3qRZs9gG+7tQ/Q3YVsGZbt+jO93cIEWj/91VaMyPLAiGSwMtLv2gF5f11qXqP8tE5Tsuqqk4Vknl4UoucxndEvSPFPYQA5rVffTDMNQNv1x7JCQiIiIoLCzEsixM08SyLE6dOkVExL/msj99+jTHjh3j0Ucdc+yfO3cOrTUXLlzg5Zdf9kSZUKlRRzU6Wl1zF6kQ7tSx42hCQjqTfekDeuy5B7uD8y9Kxlc26hRY4wyOFK9mx45XiY19gJ49H/Rcwdeh+yv0t46JBa3pyjH27uOMLBv1PViTDPffNGcqdGeF7gyMAM45hsbVMY06qKEbcBPLg9SVR0IiLCyM2NhY0tLSSEpKIi0tjdjY2BpDTZGRkXz99dfVP7/11ltcunSJ5557zhMlAqCOa1Ql2N18/x+38C5KGcTEzGDH9te4eDifwGGdoMV17mQ+ZGPs1dh9FcVtDrMx5THatRvI6NGvNEDVV/Fz3Dthptuo3dpxj4APU8dtjH2O34Xu2ABXQrZU6N4K3du9b+Oxzl588UWWLFlCQkICS5YsYcECx9UXs2bNYs+ePZ4q44bUEY0OAB3p2/+4hXeKjr4Xjc23+kOM610Oe1ZjfGaj20PpgIusXZuMn19zEhKWeM08R7qLgd1VYWy34bwPn8Quq1rMqQ3YQ337Unml6zOXrxer7zmJtm1COPtqCbq7wo7zvYVNGv3Y7w34Um8ffTSWirPnSVZfYP/cj/DIlo7eKjTmPyy4DJXTDNZ98QC5uZ8wefLHdOz444Yuu6bzGnO5he6isMfd+LPUWH93xgYLdURjTTEh/PpfKhtLb7Wdk/DtCKyDiiMVqArQ3eUoQjScmJhkistzOFO+G3Wg6suO1hibbDjrmNl156G3OHJkFcOGveR9AQHQQmEPNDCOaNTxetwg6OXUIRvjkMYeZDgNCF8iIVGlYn8FurkMNYmG1aPHVEyzOfsDlmHssdFao/ZrjIMae4jBcTaRlTWf7t2n0L//vzV0uU7pfgrdCozP7TrfIOjVLjrW6tDt8PlzLldISABYmooDFeiopnFFhvBeAQFtiIqayIHKj7BLyin9vBRjs43dWXHutuOsX/8grVtHM2bMIu+eLdV0nMRW50Dt9JGQuDIVuwXWmGvX0fBVEhLgWKK0rGqZUiEaWEzMvZRWFJHXfAOln5VCEJT/uJy1636ObVcyYcJS/P3rf9Oop+jOBnY3hbHDhnONPyhUtsY47piKndZNZ18hIQFof4VfDz90p6bzixfeq0uXeAID27E/6APwg8p4g01bn+H06W8YO/ZdWre+raFLvGn2CAMUGF828nMT32uMr2zsTgrdq2ntJyQkACIULe5tIUNNwisYhh/R0T8lr2QdzR63yC76X3Jy/sagQc/StevEhi6vbkIU9mADI0875h1qjOyqyfsMx1TsTW1ONwkJIbxQTEwytl3Bhs+e5fPPn6FLl3iGDHm+ocuqF91Hods47sSmsvENO6mdGlUI9mgDQppWQICEhBBeqW3b3oSF9WHnzr8SEtKR+Pj/wTAa6f07V05inwfjm0Z2NHFGY2yzsbsr9G1Nc3fZNLsWohHo2/dxAgLaMH78EgICQmvfwIvpjgZ2D+W40un7RnI0UVk1zBRQdRTRRDXdzoXwcrGx9/HMM4W0bdu3oUtxCXu4AQYYX9heu1TnDxlbbVRx1XmIgKY3zHSFhIQQXsw0fWhNk+Cqk9jHNCrPy0MiX6N2aeyeCt2lae8mm3b3QgiP0r0VOrTqaMKL1nGuoVxjfmpBy6qjnyZO/gaEEJ5jKqzRJuqC957ENr604ULVXdX+TXeY6QoJCSGEZ0Uo7GjHSeyK3IqGrqYGlWdj5Gh0f1W9OmBTJyEhhPA4e5gBQXBhyQWMNAsKvWDo6XLVeh1hYA+WXeMV8jchhPC8IIU1wyQwPhB1RuO30sJYa8GZBgqLK9Oxl4E11pTZF37AI8uXCiHENfwVAcMDON+lHLVHY+yy8UuxsLs7roKijed21OpbjZGrsYYZECoB8UMSEkKIhtVMoQcprN4KY5eN2qMxj1jo25RjYZ9Wbt5pn9cYX9joCNB9JSCuJiEhhPAOzRX2UBP6aIydNmqvxjxkoWOqwsId8yZpjfGpDRqsO5vOGhF1ISEhhPAugQp7uAl9NcY3NipbYx6w0D0dy6IS5LodudqjMfI1VpwBLSUgrkdCQgjhnYIV9igT+mmM7TZqn8bMsdC9FHZ/AwJvcaderDG+trF/pNAxEhDOSEgIIbxbC4UdZ8IAx4ysapfGzLbQfRV2XwOa12MHb1VN3ucP9k+a3hoRdSEhIYRoHFop7LH/Cgtju0bttbD7Geg+qk53Rxs7bNQZsBJcO3zliyQkhBCNS6jCHmdin9EYW23MLTZ6N9gDDMfSon617PQLNWqHxo5W6K5yq1htJCSEEI1TW4U9wcQu1BhbbMyvbPQusAca6Fh1/RviKqqGmYLBHikBcTPkb0kI0bi1V9iTTKxJBrQEc7ONucxC5dhg17yD28iyUd+DfWc9z2U0QXIkIYTwCbqjgRWpUCeqjiwybfQ3jnmYdHeFOqkx9mnsPgrdUb4f3ywJCSGE71AK3VlhdVKovKpzFhtt9A6gDHQbsO+QgKgLCQkhhO9RCt1VYUUp1GHH1VCUgjXBrP3EtqhBQkII4buUQvdQWN0UlHHrN+A1QRISQgjfZygIbOgiGicZnBNCCOGUhIQQQginJCSEEEI45bFzErm5ucyZM4eSkhJat27NwoULiYqKqvGcRYsWkZ6ejmEY+Pv78+STTzJ69GhPlSiEEOIqHguJ+fPnk5ycTFJSEqtWrWLevHm8//77NZ7Tt29fHn74YQIDA8nJyeG+++5j8+bNBAQEeKpMIYQQP+CR4aaioiKys7NJTEwEIDExkezsbIqLi2s8b/To0QQGOi5BiImJQWtNSUmJJ0oUQghxHR4JiYKCAtq3b49pmgCYpkm7du0oKChwuk1qaipdunShQ4cOnihRCCHEdXjlfRJbtmzhjTfe4C9/+Uudtw0LC6n3+4aHt6j3to2BL/cnvTVevtyfL/TmkZCIiIigsLAQy7IwTRPLsjh16hQRERHXPPebb77h2Wef5c9//jPdunWr83sVFV3Avmrmx5sRHt6C06fP13m7xsKX+5PeGi9f7q+x9GYY6oZfrj0y3BQWFkZsbCxpaWkApKWlERsbS2hoaI3n7d69myeffJI333yTXr16eaI0IYQQN+Cx+yRefPFFlixZQkJCAkuWLGHBggUAzJo1iz179gCwYMECSktLmTdvHklJSSQlJXHgwAFPlSiEEOIqSmtd97EZLybDTdfny/1Jb42XL/fXWHrziuEmIYQQjZOEhBBCCKckJIQQQjglISGEEMIpCQkhhBBOSUgIIYRwSkJCCCGEUxISQgghnJKQEEII4ZSEhBBCCKckJIQQQjglISGEEMIpCQkhhBBOSUgIIYRwSkJCCCGEUxISQgghnJKQEEII4ZSEhBBCCKckJIQQQjglISGEEMIpCQkhhBBOSUgIIYRwSkJCCCGEUxISQgghnJKQEEII4ZSEhBBCCKckJIQQQjglISGEEMIpCQkhhBBOSUgIIYRwSkJCCCGEUxISQgghnJKQEEII4ZSEhBBCCKckJIQQQjjlsZDIzc1lxowZJCQkMGPGDPLy8q55jmVZLFiwgPj4eO666y4+/PBDT5UnhBDiOjwWEvPnzyc5OZl169aRnJzMvHnzrnnOJ598wrFjx1i/fj3Lly/nrbfe4sSJE54qUQghxFX8PPEmRUVFZGdns3jxYgASExN5+eWXKS4uJjQ0tPp56enpTJ8+HcMwCA0NJT4+nrVr1/LII4/c9HsZhqp3nbeybWPgy/1Jb42XL/fXGHqrrUaPhERBQQHt27fHNE0ATNOkXbt2FBQU1AiJgoICIiMjq3+OiIjgu+++q9N7tWkTXO86w8JC6r1tY+DL/UlvjZcv9+cLvcmJayGEEE55JCQiIiIoLCzEsizAcYL61KlTREREXPO8/Pz86p8LCgro0KGDJ0oUQghxHR4JibCwMGJjY0lLSwMgLS2N2NjYGkNNAOPHj+fDDz/Etm2Ki4vZsGEDCQkJnihRCCHEdSittfbEGx0+fJg5c+Zw7tw5WrZsycKFC+nWrRuzZs3iiSeeoE+fPliWxUsvvcQXX3wBwKxZs5gxY4YnyhNCCHEdHgsJIYQQjY+cuBZCCOGUhIQQQginJCSEEEI4JSEhhBDCKZ8MibNnzzJr1iwSEhKYNGkSs2fPpri4GICdO3cyefJkEhISePjhhykqKqre7umnn2bUqFHExMRw8eLFGq95o+08zdX95ebmcv/99zN+/HgSExOZO3cupaWlHu8L3PO7u2Lu3Lk3fNzd3NFbSUkJTz31FAkJCdx99928/fbbHu3ph9zRX0pKCpMmTSIpKYmpU6eybds2j/Z0RX16q+1zlZGRwfjx47nrrrv4zW9+w+XLlxukt1ppH3T27FmdlZVV/fOf/vQnPXfuXG1Zlo6Pj9dbt27VWmu9aNEiPWfOnOrnffnll/rMmTM6OjpaX7hwofrPa9vO01zd3/Hjx/W+ffu01o5ef/3rX+u3337bQ93U5Orerti4caOeO3eu08c9wR29PfbYY3rx4sXVP586dcq9TdyAq/srLi7WAwYM0KdPn9Zaa71hwwY9YcIED3VTU316u9Hn6sKFC3rEiBE6NzdXa631888/r9966y0PdnTzfPJIonXr1txxxx3VP/fv35/8/Hz27t1L8+bNGTx4MAAzZ85k7dq11c8bPnw4YWFh17xebdt5mqv769SpEz179gTAMAz69u1b4853T3J1b+D4Fvj2228zd+5c9xZfC1f3lpeXx7fffssDDzxQ/Wfh4eFu7ODGXN2f1hqtdfXRxfnz5xtsBob69Hajz9WmTZvo3bs3UVFR1dutWbPGgx3dPJ8MiR+ybZtly5YxZsyYayYQDA0NxbZtSkpKbvga9d3OE1zR3w+Vlpby0UcfMWbMGHeUWyeu6u2ll17iiSeeoEWLFu4st05c0duhQ4do3749L7zwAlOmTGHWrFkcPHjQ3aXfFFf0FxoayksvvcSUKVOIi4vjtddeY/78+e4uvVb16e3qz9XV20VGRlJQUOCZBurI50Pi5ZdfJigoiPvuu6+hS3ELV/ZXWVnJk08+ybBhwxg7dqwLqrs1rugtPT0df39/4uLiXFeYC7iiN9u22bVrF1OnTmXlypVMnz6dxx9/3IVV1p8r+rtw4QJLly4lJSWFzMxM5syZw+zZs9ENfP9vXXvzts9VXfl0SCxcuJCjR4/yn//5nxiGcc0EgsXFxRiGQevWrW/4OvXdzt1c1R84Jl185plnaNWqFf/+7//uzrJviqt627JlC1lZWYwZM6b6W1xiYiKHDh1ya/034sp/lxEREdVDHePGjeP06dPVJ1SMYumIAAAE8klEQVQbiqv627x5My1atKBbt24ATJw4kWPHjnH27Fm31n8jde3N2efq6u3y8/OvmfDUW/hsSLz22mvs3buXRYsW0axZMwB69+5NaWlp9RUSH3zwAePHj6/1teq7nTu5sj/btpkzZw6mafL73/8epRp2oRRX9vbiiy+yadMmMjIyyMjIABwTTPbo0cN9DdyAq/9dBgUFVQ8xbd26lVatWtGmTRv3NVALV/bXqVMnsrOzq68WysrKIiQkpMH6q2tvN/pcjR49mj179lQv4/zBBx8wYcIEzzZ0k3xy7qaDBw+SmJhIVFQUAQEBgOMf3KJFi9ixYwfz58+nrKyMjh078sorr9C2bVsAZs+eze7duyksLKRdu3ZER0fz3nvvAdxwu8beX2ZmJo899hjR0dEYhuN7w8CBAxtk/Ncdv7sfiomJYceOHQQH139xqvpyR2979uxhwYIFlJeXExgYyAsvvEDfvn093pu7+lu8eDErVqzA39+fZs2aMWfOnOojJ2/vrbbP1YYNG3jllVewbZvY2Fj+9Kc/ERQU5PHeauOTISGEEMI1fHa4SQghxK2TkBBCCOGUhIQQQginJCSEEEI4JSEhhBDCKQkJIRrYmjVreOihhygrK2voUoS4hoSEEDfh/vvvZ8iQIZSXl7v0dbOzs0lJSWHRokU0b97cpa8thCtISAhRixMnTrBt2zaUUmzcuNGlr92zZ0/ee+89r7yJSgiQkBCiVqmpqfTr148pU6aQmppa/edz5sxhwYIFPProowwYMIDp06dz7Nix6sdjYmJYtmwZ48aNY/DgwSxYsKDG5HQpKSlMmDCBIUOG8Itf/IKTJ09WP3b48GEeeughhg4dSkJCAunp6Z5pVoirSEgIUYtVq1YxadIkJk2axObNmzlz5kz1Y+np6cyePZutW7fSpUsXXn/99RrbZmZmkpKSwscff8yaNWv4/PPPAceUDO+88w5vv/02X331FYMGDeLpp58G4NKlSzz88MMkJiby5Zdf8vrrr7NgwYIGnZRQNF0SEkLcwLZt28jPz2fChAn07t2bzp07k5aWVv14fHw8ffv2xc/Pj8mTJ7N///4a28+aNYuWLVsSGRnJHXfcQU5ODuCY0O3RRx+le/fu+Pn58ctf/pL9+/dz8uRJMjMz6dixI9OmTcPPz4+ePXuSkJDQoAtdiabLr6ELEMKbpaamMnLkSEJDQwHHNOMrV67kwQcfBKgxyWNAQACXLl2qsf0PV4oLDAysXmUtPz+fP/zhDyxcuLD6ca01hYWFnDx5kt27d9eYyM6yLCZPnuzy/oSojYSEEE6UlpayZs0abNtm5MiRAJSXl3Pu3LnqI4L6ioiI4Je//OV1d/z5+fkMGTKExYsX39J7COEKMtwkhBMbNmzANE1Wr15NamoqqamppKenM3jw4BonsOtj5syZvPvuu9VrQZw/f756jeO4uDjy8vJITU2loqKCiooKdu/ezeHDh2+5JyHqSkJCCCdWrlzJ1KlTiYyMJDw8vPq/n/3sZ3zyySdUVlbW+7XvuusuHnnkEZ566ikGDhxIYmIimzZtAiAkJIT33nuP9PR0Ro8ezahRo3j11Vddfo+GEDdD1pMQQgjhlBxJCCGEcEpCQgghhFMSEkIIIZySkBBCCOGUhIQQQginJCSEEEI4JSEhhBDCKQkJIYQQTklICCGEcOr/A2rsK2WXxDBuAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ],
      "metadata": {
        "datalore": {
          "sheet_delimiter": false
        },
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 649
        },
        "id": "vJvHOztTR0H7",
        "outputId": "b0919e46-e37c-49ae-a654-bd232c831e84"
      }
    }
  ],
  "metadata": {
    "colab": {
      "name": "mds.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "language_info": {
      "name": "python"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}