{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mkaD5bqYSc4M"
      },
      "source": [
        "🏦 RemiCash — Agente de Scoring Crediticio Alternativo\n",
        "## Entrega 2: Prototipo Funcional y Resultados Iniciales\n",
        "**Data Science & IA · Máster Fintech, Blockchain y Mercados Financieros · UB 2025–2026**\n",
        "\n",
        "Equipo: Kilian Martorell · Alison Berbetty · Ronald Rodriguez · Andrés Felipe Vera\n",
        "\n",
        "---\n",
        "### Flujo del PoC\n",
        "```\n",
        "Datos Open Banking (sintéticos)\n",
        "        ↓\n",
        "Preprocesamiento & Feature Engineering\n",
        "        ↓\n",
        "Modelo ML (XGBoost) + Métricas\n",
        "        ↓\n",
        "Explicabilidad SHAP\n",
        "        ↓\n",
        "Agente LLM (Claude API) → Decisión explicable\n",
        "```"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dSVoTUCHSc4N"
      },
      "source": [
        "## 0. Instalación de dependencias"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YR0EnpCbSc4O",
        "outputId": "cfd361e5-082b-4aec-f26e-f20ea5634acb"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Requirement already satisfied: xgboost in /usr/local/lib/python3.12/dist-packages (3.2.0)\n",
            "Requirement already satisfied: shap in /usr/local/lib/python3.12/dist-packages (0.51.0)\n",
            "Requirement already satisfied: pandas in /usr/local/lib/python3.12/dist-packages (2.2.2)\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python3.12/dist-packages (2.0.2)\n",
            "Requirement already satisfied: scikit-learn in /usr/local/lib/python3.12/dist-packages (1.6.1)\n",
            "Requirement already satisfied: matplotlib in /usr/local/lib/python3.12/dist-packages (3.10.0)\n",
            "Requirement already satisfied: seaborn in /usr/local/lib/python3.12/dist-packages (0.13.2)\n",
            "Collecting anthropic\n",
            "  Downloading anthropic-0.105.2-py3-none-any.whl.metadata (3.2 kB)\n",
            "Requirement already satisfied: nvidia-nccl-cu12 in /usr/local/lib/python3.12/dist-packages (from xgboost) (2.30.4)\n",
            "Requirement already satisfied: scipy in /usr/local/lib/python3.12/dist-packages (from xgboost) (1.16.3)\n",
            "Requirement already satisfied: tqdm>=4.27.0 in /usr/local/lib/python3.12/dist-packages (from shap) (4.67.3)\n",
            "Requirement already satisfied: packaging>20.9 in /usr/local/lib/python3.12/dist-packages (from shap) (26.2)\n",
            "Requirement already satisfied: slicer==0.0.8 in /usr/local/lib/python3.12/dist-packages (from shap) (0.0.8)\n",
            "Requirement already satisfied: numba in /usr/local/lib/python3.12/dist-packages (from shap) (0.60.0)\n",
            "Requirement already satisfied: llvmlite in /usr/local/lib/python3.12/dist-packages (from shap) (0.43.0)\n",
            "Requirement already satisfied: cloudpickle in /usr/local/lib/python3.12/dist-packages (from shap) (3.1.2)\n",
            "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.12/dist-packages (from shap) (4.15.0)\n",
            "Requirement already satisfied: python-dateutil>=2.8.2 in /usr/local/lib/python3.12/dist-packages (from pandas) (2.9.0.post0)\n",
            "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.12/dist-packages (from pandas) (2025.2)\n",
            "Requirement already satisfied: tzdata>=2022.7 in /usr/local/lib/python3.12/dist-packages (from pandas) (2026.2)\n",
            "Requirement already satisfied: joblib>=1.2.0 in /usr/local/lib/python3.12/dist-packages (from scikit-learn) (1.5.3)\n",
            "Requirement already satisfied: threadpoolctl>=3.1.0 in /usr/local/lib/python3.12/dist-packages (from scikit-learn) (3.6.0)\n",
            "Requirement already satisfied: contourpy>=1.0.1 in /usr/local/lib/python3.12/dist-packages (from matplotlib) (1.3.3)\n",
            "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.12/dist-packages (from matplotlib) (0.12.1)\n",
            "Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.12/dist-packages (from matplotlib) (4.63.0)\n",
            "Requirement already satisfied: kiwisolver>=1.3.1 in /usr/local/lib/python3.12/dist-packages (from matplotlib) (1.5.0)\n",
            "Requirement already satisfied: pillow>=8 in /usr/local/lib/python3.12/dist-packages (from matplotlib) (11.3.0)\n",
            "Requirement already satisfied: pyparsing>=2.3.1 in /usr/local/lib/python3.12/dist-packages (from matplotlib) (3.3.2)\n",
            "Requirement already satisfied: anyio<5,>=3.5.0 in /usr/local/lib/python3.12/dist-packages (from anthropic) (4.13.0)\n",
            "Requirement already satisfied: distro<2,>=1.7.0 in /usr/local/lib/python3.12/dist-packages (from anthropic) (1.9.0)\n",
            "Requirement already satisfied: docstring-parser<1,>=0.15 in /usr/local/lib/python3.12/dist-packages (from anthropic) (0.18.0)\n",
            "Requirement already satisfied: httpx<1,>=0.25.0 in /usr/local/lib/python3.12/dist-packages (from anthropic) (0.28.1)\n",
            "Requirement already satisfied: jiter<1,>=0.4.0 in /usr/local/lib/python3.12/dist-packages (from anthropic) (0.15.0)\n",
            "Requirement already satisfied: pydantic<3,>=1.9.0 in /usr/local/lib/python3.12/dist-packages (from anthropic) (2.12.3)\n",
            "Requirement already satisfied: sniffio in /usr/local/lib/python3.12/dist-packages (from anthropic) (1.3.1)\n",
            "Requirement already satisfied: idna>=2.8 in /usr/local/lib/python3.12/dist-packages (from anyio<5,>=3.5.0->anthropic) (3.15)\n",
            "Requirement already satisfied: certifi in /usr/local/lib/python3.12/dist-packages (from httpx<1,>=0.25.0->anthropic) (2026.5.20)\n",
            "Requirement already satisfied: httpcore==1.* in /usr/local/lib/python3.12/dist-packages (from httpx<1,>=0.25.0->anthropic) (1.0.9)\n",
            "Requirement already satisfied: h11>=0.16 in /usr/local/lib/python3.12/dist-packages (from httpcore==1.*->httpx<1,>=0.25.0->anthropic) (0.16.0)\n",
            "Requirement already satisfied: annotated-types>=0.6.0 in /usr/local/lib/python3.12/dist-packages (from pydantic<3,>=1.9.0->anthropic) (0.7.0)\n",
            "Requirement already satisfied: pydantic-core==2.41.4 in /usr/local/lib/python3.12/dist-packages (from pydantic<3,>=1.9.0->anthropic) (2.41.4)\n",
            "Requirement already satisfied: typing-inspection>=0.4.2 in /usr/local/lib/python3.12/dist-packages (from pydantic<3,>=1.9.0->anthropic) (0.4.2)\n",
            "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.12/dist-packages (from python-dateutil>=2.8.2->pandas) (1.17.0)\n",
            "Downloading anthropic-0.105.2-py3-none-any.whl (837 kB)\n",
            "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m837.5/837.5 kB\u001b[0m \u001b[31m11.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hInstalling collected packages: anthropic\n",
            "Successfully installed anthropic-0.105.2\n"
          ]
        }
      ],
      "source": [
        "# Ejecutar solo si no están instaladas\n",
        "!pip install xgboost shap pandas numpy scikit-learn matplotlib seaborn anthropic"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "or4pUU6gSc4O"
      },
      "source": [
        "## 1. Importaciones"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "j638L6ElSc4O",
        "outputId": "817c106a-e3e7-4f57-c829-c852ea13c2e8"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "✅ Librerías cargadas correctamente\n"
          ]
        }
      ],
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import matplotlib.patches as mpatches\n",
        "import seaborn as sns\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')\n",
        "\n",
        "from sklearn.model_selection import train_test_split, StratifiedKFold, cross_val_score\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.metrics import (\n",
        "    roc_auc_score, classification_report, confusion_matrix,\n",
        "    f1_score, precision_score, recall_score, roc_curve\n",
        ")\n",
        "from xgboost import XGBClassifier\n",
        "import shap\n",
        "import anthropic\n",
        "import json\n",
        "\n",
        "# Seed para reproducibilidad\n",
        "SEED = 42\n",
        "np.random.seed(SEED)\n",
        "\n",
        "print('✅ Librerías cargadas correctamente')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0ILSuxDySc4P"
      },
      "source": [
        "## 2. Generación del Dataset Sintético — Open Banking Simulado\n",
        "\n",
        "Simulamos perfiles de migrantes latinoamericanos en España con variables de comportamiento financiero accesibles vía PSD2/Open Banking. No se usan datos reales (cumplimiento GDPR)."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 258
        },
        "id": "xYkc5MelSc4P",
        "outputId": "79eb322c-7c87-4e7e-d362-547b63f3ea20"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Dataset generado: 1000 perfiles, 10 features\n",
            "Tasa de impago en dataset: 31.3%\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   ingreso_mensual_eur  antiguedad_cuenta_meses  ratio_gasto_ingreso  \\\n",
              "0              2000.00                       16                  0.7   \n",
              "1              1555.04                       16                  0.3   \n",
              "2              1575.03                       29                  0.3   \n",
              "3              1776.17                       52                  0.3   \n",
              "4              2545.54                       52                  0.3   \n",
              "\n",
              "   frecuencia_remesas_anual  regularidad_ingresos  descubiertos_12m  \\\n",
              "0                         0                0.1197                 2   \n",
              "1                         9                0.8942                 0   \n",
              "2                         8                0.6500                 0   \n",
              "3                         8                0.8550                 0   \n",
              "4                         2                0.9175                 1   \n",
              "\n",
              "   variabilidad_ingresos_eur  num_cuentas_bancarias  cuotas_deuda_mensual_eur  \\\n",
              "0                     100.00                      1                    234.63   \n",
              "1                      20.37                      2                    148.38   \n",
              "2                       0.19                      3                    161.39   \n",
              "3                      72.55                      1                    201.55   \n",
              "4                      88.78                      3                     56.10   \n",
              "\n",
              "   ahorro_medio_mensual_eur  impago  \n",
              "0                    365.37       1  \n",
              "1                    940.14       0  \n",
              "2                    941.14       0  \n",
              "3                   1041.77       0  \n",
              "4                   1500.00       0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-896b627b-a37d-4f9d-b9e7-74f72460b549\" 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>ingreso_mensual_eur</th>\n",
              "      <th>antiguedad_cuenta_meses</th>\n",
              "      <th>ratio_gasto_ingreso</th>\n",
              "      <th>frecuencia_remesas_anual</th>\n",
              "      <th>regularidad_ingresos</th>\n",
              "      <th>descubiertos_12m</th>\n",
              "      <th>variabilidad_ingresos_eur</th>\n",
              "      <th>num_cuentas_bancarias</th>\n",
              "      <th>cuotas_deuda_mensual_eur</th>\n",
              "      <th>ahorro_medio_mensual_eur</th>\n",
              "      <th>impago</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2000.00</td>\n",
              "      <td>16</td>\n",
              "      <td>0.7</td>\n",
              "      <td>0</td>\n",
              "      <td>0.1197</td>\n",
              "      <td>2</td>\n",
              "      <td>100.00</td>\n",
              "      <td>1</td>\n",
              "      <td>234.63</td>\n",
              "      <td>365.37</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1555.04</td>\n",
              "      <td>16</td>\n",
              "      <td>0.3</td>\n",
              "      <td>9</td>\n",
              "      <td>0.8942</td>\n",
              "      <td>0</td>\n",
              "      <td>20.37</td>\n",
              "      <td>2</td>\n",
              "      <td>148.38</td>\n",
              "      <td>940.14</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1575.03</td>\n",
              "      <td>29</td>\n",
              "      <td>0.3</td>\n",
              "      <td>8</td>\n",
              "      <td>0.6500</td>\n",
              "      <td>0</td>\n",
              "      <td>0.19</td>\n",
              "      <td>3</td>\n",
              "      <td>161.39</td>\n",
              "      <td>941.14</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1776.17</td>\n",
              "      <td>52</td>\n",
              "      <td>0.3</td>\n",
              "      <td>8</td>\n",
              "      <td>0.8550</td>\n",
              "      <td>0</td>\n",
              "      <td>72.55</td>\n",
              "      <td>1</td>\n",
              "      <td>201.55</td>\n",
              "      <td>1041.77</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2545.54</td>\n",
              "      <td>52</td>\n",
              "      <td>0.3</td>\n",
              "      <td>2</td>\n",
              "      <td>0.9175</td>\n",
              "      <td>1</td>\n",
              "      <td>88.78</td>\n",
              "      <td>3</td>\n",
              "      <td>56.10</td>\n",
              "      <td>1500.00</td>\n",
              "      <td>0</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-896b627b-a37d-4f9d-b9e7-74f72460b549')\"\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-896b627b-a37d-4f9d-b9e7-74f72460b549 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-896b627b-a37d-4f9d-b9e7-74f72460b549');\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>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "df",
              "summary": "{\n  \"name\": \"df\",\n  \"rows\": 1000,\n  \"fields\": [\n    {\n      \"column\": \"ingreso_mensual_eur\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 548.1798891834826,\n        \"min\": 700.0,\n        \"max\": 3165.54,\n        \"num_unique_values\": 949,\n        \"samples\": [\n          2133.72,\n          1499.34,\n          1359.49\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"antiguedad_cuenta_meses\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 16,\n        \"min\": 1,\n        \"max\": 59,\n        \"num_unique_values\": 59,\n        \"samples\": [\n          16,\n          50,\n          5\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"ratio_gasto_ingreso\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.21172938891906895,\n        \"min\": 0.3,\n        \"max\": 0.97,\n        \"num_unique_values\": 459,\n        \"samples\": [\n          0.7788,\n          0.8179,\n          0.7866\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"frecuencia_remesas_anual\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 3,\n        \"min\": 0,\n        \"max\": 11,\n        \"num_unique_values\": 12,\n        \"samples\": [\n          10,\n          11,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"regularidad_ingresos\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.25666904123350753,\n        \"min\": 0.1,\n        \"max\": 0.9852,\n        \"num_unique_values\": 724,\n        \"samples\": [\n          0.4616,\n          0.7814,\n          0.9138\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"descubiertos_12m\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 2,\n        \"min\": 0,\n        \"max\": 7,\n        \"num_unique_values\": 8,\n        \"samples\": [\n          0,\n          4,\n          2\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"variabilidad_ingresos_eur\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 176.9627556331188,\n        \"min\": 0.03,\n        \"max\": 900.0,\n        \"num_unique_values\": 860,\n        \"samples\": [\n          50.56,\n          27.39,\n          131.56\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"num_cuentas_bancarias\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 1,\n        \"max\": 3,\n        \"num_unique_values\": 3,\n        \"samples\": [\n          1,\n          2,\n          3\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"cuotas_deuda_mensual_eur\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 120.22705525875313,\n        \"min\": 0.0,\n        \"max\": 600.0,\n        \"num_unique_values\": 891,\n        \"samples\": [\n          151.13,\n          447.82,\n          250.84\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"ahorro_medio_mensual_eur\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 561.8551388561438,\n        \"min\": 0.0,\n        \"max\": 1500.0,\n        \"num_unique_values\": 709,\n        \"samples\": [\n          1467.09,\n          1375.58,\n          789.52\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"impago\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 0,\n        \"max\": 1,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          0,\n          1\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 3
        }
      ],
      "source": [
        "def generar_dataset_openbanking(n=1000, seed=42):\n",
        "    rng = np.random.default_rng(seed)\n",
        "\n",
        "    # Grupos de riesgo: 70% bajo riesgo, 30% alto riesgo\n",
        "    n_bajo = int(n * 0.70)\n",
        "    n_alto = n - n_bajo\n",
        "\n",
        "    # --- Perfil BAJO RIESGO (buenos pagadores) ---\n",
        "    ingreso_bajo     = rng.normal(2000, 400, n_bajo).clip(1200, 4000)\n",
        "    antiguedad_bajo  = rng.integers(12, 60, n_bajo)\n",
        "    ratio_bajo       = rng.beta(2, 5, n_bajo).clip(0.30, 0.65)\n",
        "    remesas_bajo     = rng.integers(2, 12, n_bajo)\n",
        "    regular_bajo     = rng.beta(6, 2, n_bajo).clip(0.65, 1.0)\n",
        "    desc_bajo        = rng.integers(0, 2, n_bajo)\n",
        "    variab_bajo      = rng.exponential(80, n_bajo).clip(0, 250)\n",
        "    cuentas_bajo     = rng.integers(1, 4, n_bajo)\n",
        "    cuotas_bajo      = rng.normal(100, 80, n_bajo).clip(0, 350)\n",
        "\n",
        "    # --- Perfil ALTO RIESGO (malos pagadores) ---\n",
        "    ingreso_alto     = rng.normal(1100, 300, n_alto).clip(700, 2000)\n",
        "    antiguedad_alto  = rng.integers(1, 18, n_alto)\n",
        "    ratio_alto       = rng.beta(5, 2, n_alto).clip(0.70, 0.97)\n",
        "    remesas_alto     = rng.integers(0, 4, n_alto)\n",
        "    regular_alto     = rng.beta(2, 5, n_alto).clip(0.10, 0.50)\n",
        "    desc_alto        = rng.integers(2, 8, n_alto)\n",
        "    variab_alto      = rng.exponential(300, n_alto).clip(100, 900)\n",
        "    cuentas_alto     = rng.integers(1, 3, n_alto)\n",
        "    cuotas_alto      = rng.normal(300, 100, n_alto).clip(150, 600)\n",
        "\n",
        "    # Concatenar\n",
        "    ingreso     = np.concatenate([ingreso_bajo, ingreso_alto])\n",
        "    antiguedad  = np.concatenate([antiguedad_bajo, antiguedad_alto])\n",
        "    ratio       = np.concatenate([ratio_bajo, ratio_alto])\n",
        "    remesas     = np.concatenate([remesas_bajo, remesas_alto])\n",
        "    regular     = np.concatenate([regular_bajo, regular_alto])\n",
        "    desc        = np.concatenate([desc_bajo, desc_alto])\n",
        "    variab      = np.concatenate([variab_bajo, variab_alto])\n",
        "    cuentas     = np.concatenate([cuentas_bajo, cuentas_alto])\n",
        "    cuotas      = np.concatenate([cuotas_bajo, cuotas_alto])\n",
        "    ahorro      = (ingreso * (1 - ratio) - cuotas).clip(0, 1500)\n",
        "\n",
        "    # Variable objetivo con muy poco ruido\n",
        "    impago_base = np.concatenate([\n",
        "        np.zeros(n_bajo, dtype=int),\n",
        "        np.ones(n_alto, dtype=int)\n",
        "    ])\n",
        "    # Pequeño ruido realista (5%)\n",
        "    flip = rng.random(n) < 0.05\n",
        "    impago = np.where(flip, 1 - impago_base, impago_base)\n",
        "\n",
        "    # Shuffle para mezclar los grupos\n",
        "    idx = rng.permutation(n)\n",
        "\n",
        "    df = pd.DataFrame({\n",
        "        'ingreso_mensual_eur':       ingreso[idx].round(2),\n",
        "        'antiguedad_cuenta_meses':   antiguedad[idx],\n",
        "        'ratio_gasto_ingreso':       ratio[idx].round(4),\n",
        "        'frecuencia_remesas_anual':  remesas[idx],\n",
        "        'regularidad_ingresos':      regular[idx].round(4),\n",
        "        'descubiertos_12m':          desc[idx],\n",
        "        'variabilidad_ingresos_eur': variab[idx].round(2),\n",
        "        'num_cuentas_bancarias':     cuentas[idx],\n",
        "        'cuotas_deuda_mensual_eur':  cuotas[idx].round(2),\n",
        "        'ahorro_medio_mensual_eur':  ahorro[idx].round(2),\n",
        "        'impago':                    impago[idx]\n",
        "    })\n",
        "    return df\n",
        "\n",
        "df = generar_dataset_openbanking(n=1000)\n",
        "print(f'Dataset generado: {df.shape[0]} perfiles, {df.shape[1]-1} features')\n",
        "print(f'Tasa de impago en dataset: {df[\"impago\"].mean():.1%}')\n",
        "df.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 447
        },
        "id": "loot24XWSc4P",
        "outputId": "505d4b7e-973a-4102-cc23-c6ca67c36709"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1200x400 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABDYAAAGdCAYAAADpMPPGAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjAsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvlHJYcgAAAAlwSFlzAAAPYQAAD2EBqD+naQAAqptJREFUeJzs3Xd8zdf/B/DXTXJv9pZJhkgQI0qsmEHsVRRF7WqrRq1WqVotqjWKqpYiaFE1Y++RWrVijyARRIzsIfOe3x9++XxzZS+fjNfz8chD7vmczznvz+feG5/7vudzjkIIIUBEREREREREVAppyR0AEREREREREVFBMbFBRERERERERKUWExtEREREREREVGoxsUFEREREREREpRYTG0RERERERERUajGxQURERERERESlFhMbRERERERERFRqMbFBRERERERERKUWExtEREREREREVGoxsUFERGXa/fv3YWpqCh8fH6SkpMgdDhEREREVMSY2iOidGTJkCBQKBYKDg+UOJVulIUbKu8TERPTu3RuVK1fG9u3boVQqM9VxdnaGs7OzRpmvry8UCgV8fX0L1K+3tzcUCkWB9i1Nystx5mTmzJlQKBQ4ceLEO++7KM5/YV/rREREJQETG0SlWHBwMBQKhcaPUqlExYoV0adPH1y8eFHuEPPl0qVLGD58ONzc3GBoaAh9fX1UqVIFAwcOxOHDh+UOL9/efm50dHRgZ2eH999/H6dOnSqSPkrrORNCwNXVFQqFAp07dy62fsaOHYuIiAjs27cPJiYmxdbPu+Ls7JzpdfX2T3lPyvGDevFKT/5m/Ltmbm6OGjVqYMCAAdi6dSuSk5OLpC+FQgFvb+8iaau48PVGRFQy6MgdABEVXpUqVfDRRx8BAOLj43Hp0iX8888/2LlzJ44cOYIWLVrIHOEb8+bNw9dff42KFStqlKvVakyaNAmLFy+Gjo4OWrdujW7dukGpVOLhw4fYu3cv/vzzT8yePRvffvutTNEXjKWlJUaPHg3gzeiBgIAA7Nq1C35+fvj777/Ru3fvArVb2s/ZiRMn8ODBAygUChw8eBChoaGwt7cv0j5evnyJihUr4uDBg/luu0ePHmjcuDHs7OwK1Pf69euRkJBQoH1zo62tjWnTpmW73czMrFj6payNHj0aH374IRwdHeUO5Z0aPnw4KlWqBCEEYmJiEBgYiN27d2Pjxo1wd3fH5s2b4eHhIXeYRERUTjCxQVQGuLq6YubMmRplP/zwA6ZMmYJvv/0WJ0+elCewt9jZ2WX5QXHatGlYvHgx3nvvPWzduhVVqlTR2P769Wv88ssvCA8Pf1ehFpkKFSpkem7++OMPjBgxAl999VWBExul/ZytXr0aADBx4kQsWLAAvr6+mDp1apH2YWVlhRkzZhRoX1NTU5iamha47+L8kKujo5PpNUXyqVChAipUqCB3GO/cxx9/jMaNG2uUxcbGYsaMGVi8eDHatWuHy5cvF3nCkoiIKEuCiEqtoKAgAUC0b98+07YXL14IAMLQ0DDTtqSkJLFw4UJRt25dYWBgIIyMjESzZs3Erl27MtUdPHiwACAePHggfvrpJ+Hm5ib09PSEu7u72LRpk9Te1KlThZOTk9DV1RW1a9cW+/bty7atoKAgqSwwMFBoa2sLS0tLERYWluPxJiYmSr/fvXtXfPnll6Ju3brCwsJC6OrqCjc3NzF58mQRGxubad/Q0FAxduxY4erqKvT09ISpqamoXr26+PTTT0VUVFSmGB8+fCiWLFkiqlWrJlQqlXB0dBQzZ84UaWlpOcaYEQBRrVq1TOVpaWnC0NBQABAvXryQytesWSMaNmwoDA0NhaGhoWjYsKFYu3Ztpv0Les5KisjISKGnpydq1aolEhIShLGxsahSpYpQq9WZ6qa/xgcPHiwCAwPF+++/L8zMzISBgYFo06aNCAgIyLTPsWPHxNChQ0XVqlWlc+np6Sl+//33LONxcnISTk5OGmVr164VADKd/0uXLolevXoJBwcHoVKpRIUKFUT9+vXF999/r1GvZcuW4u3/YtPS0sSqVatEgwYNhLm5udDT0xMVK1YUXbp0EcePH8/9xP1/rLq6unmq+++//wptbW1Rp06dTK+DrLYdP35cABAzZswQ/v7+omXLlsLIyEiYmpqKnj17isDAwEx9ZHWcQgiRkpIiFi5cKDw8PISenp4wMTER3t7ews/PL1PdjOfaz89PNGnSRBgZGUnPSVJSkli6dKlo166dqFSpklCpVMLKykr06NFDXL58WaOt9PdvVj8ZBQcHi2HDhgl7e3uhVCpFxYoVxbBhw8SjR4+yPcbXr1+Lb775Rri4uAgdHR0xY8YMIYQQM2bMEACyfA79/PyEt7e3MDExEXp6esLDw0MsXLhQpKSkZKqbE39/f9GiRQthYGAgLCwsRJ8+fURISEi251+tVovVq1eLJk2aCGNjY6Gvry88PT3F6tWrM9XN7rWenfRzfPbs2WzrDBkyRAAQI0eO1CjP63sz/bWY1U96nFFRUeKHH34QLVq0EHZ2dkKpVAo7OzsxcOBAcf/+/UwxvX79WixYsEB4eHgIExMTYWBgIJycnETv3r2z/Duyc+dO0bp1a2FmZiZ0dXVFzZo1xU8//SRSU1MznYvcXm9ERFT8OGKDqIzT0dF8myclJaFDhw44ceIE3nvvPQwfPhwpKSnYu3cvunfvjmXLlkm3TmQ0YcIEnD9/Hl27doW2tjY2b96M/v37w9zcHMuWLcOtW7fQuXNnJCYmYuPGjejevTtu376daSTB23x9fZGWloZPP/0UNjY2OdbV1dWVft++fTtWr16NVq1awdvbG2q1GufOncP8+fNx8uRJnDp1SpooMiEhAU2bNkVwcDDatWuHHj16IDk5GUFBQdiwYQMmTZqU6dv5L7/8EidPnkSXLl3Qvn177Ny5EzNnzkRycjLmzJmTY5z5kT7x39ixY7Fs2TJUrFgRw4cPBwBs27YNQ4cOxZUrV7BkyRJpn4Kes5Ji48aNSExMxKBBg6Cvr48PPvgAa9euxcmTJ7O9nz44OBiNGzdGzZo1MWzYMDx48AC7du1Cq1atcPv2bY3zMH/+fDx8+BANGzZEjx49EB0djYMHD+LTTz/F3bt3sXDhwgLFHRAQgCZNmkBbWxvdu3eHk5MToqKicOvWLaxcuRLffPNNjvtPmTIFP/74I6pUqYL+/fvD2NgYT58+xb///osjR44U+VwCTZs2xbRp0zBr1ixMnjwZP//8MwAgKioKAwYMgK6uLjZt2pTpNXLu3DnMmzcPHTp0wJgxY3Dz5k3s2LED/v7+OHfuHFxcXHLsVwiBDz74ALt27ULVqlUxatQoxMfH4++//0a3bt2waNEijB8/PtN+//zzDw4dOoQuXbrg888/R0xMDAAgIiIC48aNQ/PmzdGpUyeYm5vj4cOH8PPzw/79+3Hq1Ck0aNAAAPD+++8jKioKu3btQvfu3fHee+9l6ufevXto1qwZXr58ia5du6JmzZq4ceMG1qxZg927d+Pff/9F1apVM+3Xq1cvXL16FR06dICZmRkqV66c43lYtGgRJk6cCAsLC/Tv3x+Ghobw8/PDxIkT4e/vj+3bt+dp4s+jR4+iY8eO0NLSQt++fWFvb4+jR4+iadOmMDc3z/L8DxgwAJs2bYKbmxv69+8PlUqFw4cPY/jw4bh16xYWLFiQa7+F8e2338LX1xdbtmzB8uXLpeOcP38+7t+/j8aNG6NHjx6IiorCgQMHMr03nZ2dMWPGDMyaNQtOTk4YMmSI1Hb6c3r79m1Mnz4drVq1Qo8ePWBoaIg7d+5g48aN2Lt3Ly5fvgwnJydpv8GDB2PLli3w8PDA0KFDoauri8ePH+P48eO4cOEC6tSpI9WdMmUKfvjhB1SsWBE9e/aEqakp/P398eWXX+L8+fP4559/AOTt9UZERO+I3JkVIiq4nEZszJ07VwAQnTt31iifOnWqACC+/fZbjW/IY2JiRP369YVKpRJPnz6VytO/kapatarG6ILz588LAMLMzEw0a9ZMxMXFSdv+/vtvAUCMGTNGo++sRmx4e3sLAOLIkSP5OvYnT56IpKSkTOWzZs0SAMSff/4plfn5+QkAYty4cZnqx8bGanybnR5j5cqVRWhoqFT+8uVLYWZmJoyNjbPsNyvIZsTGmjVrpD6EEOLkyZMCgHB3d9cYPRIRESGqVq0qAIhTp05J5QU9ZyVFvXr1hJaWlvQ6O3bsmAAgPvroo0x101/jAMQPP/ygsW3atGkCgJg3b55G+ePHjzO1k5qaKjp37iy0tbUzfSuf1xEbEyZMEADEzp07M7X/6tUrjcdZfZNuYWEh7O3tRXx8fKb9w8PDM5VlxcnJSWhra4sZM2Zk+bNixQqN+qmpqaJp06ZCoVBIo6j69OkjAOT4Lflvv/2mse23334TAESXLl1yPc5169YJAKJly5Ya75VHjx6JChUqCB0dHfHgwQOpPP1ca2lpicOHD2c65sTERPHkyZNM5Tdu3BBGRkbCx8dHozy3EQitWrXK8viXL18uAIjWrVtneYzvvfdels9TViM27t+/L3R0dIS1tbUICQnROJZmzZoJAGL9+vVZxpdRWlqacHFxEQqFQvj7+0vlarVa9O/fP8vRAStXrhQAxNChQ0VycrJUnpSUJLp27SoAiIsXL0rlxTFiQwghHBwcpNF+6R4+fJipXkpKimjbtm2W783011FWoqKisnw+jh07JrS0tMTHH3+sUVehUAhPT0+NERdCvHmPREZGSo8PHTok/b+a8f81tVotPvvsMwFAbN26VSrP7/kjIqLiwcQGUSmW/qGvSpUq0gebSZMmSRfuNjY24tatW1L9tLQ0YW5unu2w//QEwLJly6Sy9IvYdevWZarv4uIiAIiTJ09qlKempgqlUilatGihUZ5VYqN69eoCgLhz505BT4OG8PBwAUAMGTIk03FNmTIl1/3TY1yzZk22265du5anWAAIS0tL6bmZPHmy6NChg/QhLv3ieNiwYQKA+PvvvzO18ddffwkAYtiwYVJZUZ+zd+nKlSsCgGjbtq1UplarhaOjo9DX19dI7Ajxv9d45cqVM90GlL6tZ8+e2fYXGxsrwsLCxLNnz6SEkq+vr0ad/CY2Dh48mOtxZpfYcHZ2LtTtQU5OTtkOfQcg6tSpk2mf4OBgYWZmJqytraWEZ1bnLD2xUbVq1UznOi0tTbi5uQmFQqGR4MzqOFu3bi0AiPPnz2fqY86cOQKAmD17tlSWfq579OiR39MhunbtKlQqlcYH+Jw+aD569EgAEDVq1Mj0NzAtLU16b2VMRqQfY1a36gmRdWJj9uzZAoCYP39+pvqnT5/OMoGSlfSkZ9euXTNtCw4OFtra2pnOv4eHhzA0NBQJCQmZ9rl27ZoAICZOnCiVFVdio1GjRtm+Dt62bdu2LN+bOSU2clK7dm3h7OwsPY6OjhYARNOmTbP8vy+jbt26CQBZ3paUniDp1auXVMbEBhFRycBbUYjKgAcPHmDWrFkaZba2tvD394erq6tUdvfuXURGRsLe3j5TfeDNKhIAcOfOnUzbshpia2dnh4cPH2bapq2tDWtra4SGhhbgaPJGCIG1a9fC19cXN27cQHR0NNRqtbQ9Y98tWrSAnZ0dfvjhB1y9ehVdunRBy5Yt4e7unu1QcE9Pz0xllSpVAvBmKH9ehYeHS+daW1sbFSpUQPfu3TFx4kQ0b94cAHDlyhUAyPJWhFatWgF4cxtEcYiKipJuUSgMb2/vPN1K8ccffwAABg0aJJUpFAp89NFHmDt3LjZu3IiRI0dm2u+9996DlpbmCuXZPR8vX77E9OnTsWPHDjx//jxTWwV9Xfbp0wc///wzevTogb59+6Jt27Zo0aJFplV+svPhhx/i119/Ra1atfDhhx+iVatW8PLygr6+fr7i0NXVRWJiYp7rOzk54bfffsOHH36IqVOnolKlSli1alW29Zs2bZrpXGtpaaFp06YIDAzE1atX4ePjk+3+V65cgYGBARo2bJhpW06v56zqpwsICMCPP/6If//9F2FhYUhJSdHY/urVqzytYJPeb8uWLTO997W0tNCiRQvcuXMHAQEBcHBwyHN8b8vpPe3l5QU9Pb08vaevXr0KANLfioycnJzg4OCgsbxvQkICrl+/Dnt7e8yfPz/TPunnLau/8e9CbGwsFixYgJ07d+LBgweIj4/X2J7f9+aJEyfw888/4/z583j16hVSU1OlbSqVSvrdxMQEnTp1wr59+1CvXj307t0b3t7eaNCggXTLYrpz587B0NAQa9asybJPfX192c4fERFlj4kNojKgffv2OHDgAIA3H+rWrVuHyZMno1u3bvjvv/9gZGQE4M296gBw8+ZN3Lx5M9v23r7YBN5cGL4tff6O7La9/eEjK7a2trhz5w6ePn2KatWq5Vo/3dixY/HLL7/AwcEB3bp1g52dnTRXwKxZs5CUlCTVNTU1xblz5zB9+nTs3r0b+/btAwA4ODjg66+/xueff56p/ZyONy0tLc9xVqtWLdeL4JiYGGhpacHKyirTNhsbGygUCmm+AaDg5ywrUVFRWSa5CiK3xEZiYiL++usvGBkZoWfPnhrbBg0ahLlz52LNmjVZJjby+ny8fv0aLVq0wL179zB06FA0a9YMFSpUgI6ODq5evYqvv/5a47WRH40aNcKJEyekBMzatWsBAA0aNMD8+fOlD+3ZWbJkCSpXroy1a9fi+++/x/fffw89PT306dMHCxcuLNaVNdq0aQMTExPExMSgf//+sLCwyLZudvO2pJdHR0fn2FdMTEympEC69ORDxtdzbv2eOXMGrVu3BgC0a9cObm5uMDIygkKhwM6dO3H16tU8P6fp/WbXV0Hiy28/CoUCNjY2ePr0aa7tpJ9ra2vrLLfb2NhoJDYiIyMhhMDTp09zfF9n9Te+qKUnKdL/riUnJ8Pb2xuXL19G3bp1MXDgQFhaWkJHRwfBwcFYt25dvt6b//zzD/r27QsjIyO0b98ezs7OMDAwgEKhgK+vLx49epSpfvp7N30+HBMTEwwdOhRz586FgYEBgDf/T6ampsp+/oiIKH+Y2CAqY6ysrDBp0iRER0fj+++/x7Rp06Rv5NM/HPbq1Qtbt26VMcr/adq0KU6cOIGjR49KH15y8+LFCyxfvhweHh44e/asdEEKAGFhYVlekDo6OsLX1xdqtRrXrl3DoUOHsHTpUowaNQrm5ubo169fkR1TfpmYmECtVuPly5eZPsC8ePECQgiND/YFOWfZcXZ2hhCiUG3k1fbt26XRFYaGhlnWuXjxIq5duwYPD48C9bFz507cuXMH06ZNw3fffaex7cWLFwVqM6PmzZtj//79eP36Nc6fP4/du3fj119/RefOnXHjxo0cJ9bU0dHBpEmTMGnSJISGhuLkyZNYu3Yt1q9fj7CwMBw8eLDQ8WVn2LBhiImJgaWlJX7++Wf069cv24kOsxrlkrE8t2VwTUxMsj3XYWFhUp23ZTd6as6cOUhKSoK/vz+aNWumse3cuXPSqIa8SO83u2MsSHy59ZNxAkvgzWiz58+fZ9nH29LPdXbn8+3jSG/T09MTFy9ezHO8Re3hw4d4/PgxrKys4OzsDADYtWsXLl++jOHDh0sjt9Jt3rwZ69aty1cfM2fOhJ6eHi5dugQ3N7dM7b3NwMBASigGBQXh+PHj+O2337BkyRK8fv0av//+O4A351ChUODVq1f5ioeIiOSllXsVIiqNpk6dCnt7e/z666/SN3ru7u4wMTHBxYsX8zSa4l0YMmQItLW1sXLlSulWmOykf5v38OFDCCHg4+OjkdQAAH9//xzb0NLSwnvvvYevvvoKmzZtAgD4+fkV4ggKr27dugDeDKt+W3pZxg+hBTlnJcHq1asBAL1798bw4cMz/bRv316jXkGkf0vbuHHjTNuOHz9e4Hbfpq+vD29vbyxcuBBTp07F69evcfjw4Tzvb29vj379+uHAgQNwdXXFkSNH8Pr16yKLL6Ply5dj9+7d+Oijj3Do0CEAQL9+/ZCQkJBl/dOnT2vc1gUAarUaZ86cgUKh0Fg9Iit169ZFQkIC/vvvv0zbsno95+bBgwewsLDIlNRISEjA5cuXM9XX1tYGkPXIqvR+T506lSmhJ4TAqVOn8h1fVnJ6T58/fx6JiYl56iP9XGf1d+3Ro0d4/PixRpmxsTHc3d1x+/btfN0yV9TSk4p9+/aVEkIPHjwAAHTv3j1T/ez+bmtpaWU7Qu7Bgwdwd3fPlNR49uwZHj58mGN8lStXxrBhw3Dy5EkYGRlp/B/QqFEjhIeHIzAwMMc20uX0eiMioneHiQ2iMkpfXx+TJ09GSkqKdJGpo6ODkSNH4tGjR5g0aVKWyY0bN24UyTfbeeXq6oqvvvoKr169QseOHREUFJSpTmJiIhYtWoSZM2cCgPQN6JkzZzQ+gD158gRTpkzJtP/Nmzez/IY2vUxPT68oDqXABg8eDODNLTQZh8BHR0dLo0/S6wAFO2dyS/+G1NnZGX///Tf++OOPTD9///039PX18eeffxY4IZO+BOfbSYZTp05hw4YNhTqGs2fPZjm3RV5eR0lJSThz5kym8vj4eMTFxUGpVGaa16Io3LhxA5MmTYKLiwt+/fVX1KtXD3PmzMGdO3cwbty4LPe5d+9epjk4Vq1ahXv37qFz585Z3jKVUfprdcqUKRp/Yx4/foxFixZBR0cHAwYMyPMxODk5ITIyUuP2ubS0NEyaNCnLxF76bTZvf+gH3ozcatWqFW7evJlpDoWVK1fi9u3baN26dba30uRV//79oaOjg0WLFmnMG5GcnIzJkycDgMYSptlp1qwZKleujD179uDff/+VyoUQmDp1apYfpseOHYuEhASMGDEiy1smgoKCNG5fKUpxcXGYOHEifH19YWdnh6lTp0rb0v9uZzwOADh58mS2c75YWFjgyZMnWW5zcnLC/fv3Nf62JyYmYuTIkZn+b3v58iVu3LiRqY3IyEgkJSVpvHfHjh0L4M0op/Dw8Ez7hIWF4fbt2xoxAlm/3oiI6N3hrShEZdgnn3yC+fPnY/369Zg6dSqqVKmCWbNm4fLly1i6dCn27t2LFi1awNraGk+fPsX169dx9epVnD17Ntt7uovD999/j8TERCxevBjVqlVD69atUatWLSiVSgQFBeHIkSMIDw/H999/D+DNffC9evXCtm3bUL9+fbRp0wbPnz/Hnj170KZNG+mbwXSHDx/Gl19+iaZNm6Jq1aqwtLTEw4cP4efnBz09PYwaNeqdHWtWWrRogTFjxmDZsmWoVasWevXqBSEEtm3bhidPnmDs2LFo0aKFxj75PWdyW7NmDYQQGDx4cLZD+k1NTdGjRw9s3LgRO3fuRN++ffPdT5cuXeDi4oIlS5bgzp07qFu3LoKCgrBt2zb07t07yyHqeTV//nwcP34cLVq0QOXKlaGnp4fLly/j6NGjcHFxQY8ePbLd9/Xr19Lrz9PTE46OjoiLi8OePXsQFhaGSZMmSXPE5CY1NTXHhNWHH36I6tWrIzExEf369UNqaio2btwIY2NjAMDEiRNx6NAhrFq1Cu3bt0evXr009m/fvj3Gjh2Lffv2oWbNmrh58yZ2796NChUqYMmSJbnGN3DgQGzfvh27du2Ch4cHunTpgvj4ePz999+IiIjAwoULc7xl521jxozBoUOH0KxZM/Tp0wd6eno4ceIEnj59Cm9v70yjItInZP35558RGRkpJWKmTZsGAFixYgWaNWuGESNGYPfu3ahRowZu3rwJPz8/WFlZYcWKFXmOLTtVqlTB/PnzMXHiRHh4eKBPnz4wNDTE7t27cffuXXTv3h0fffRRru1oaWlh5cqV6NSpE3x8fNC3b1/Y29vj2LFjePbsGTw8PHDt2jWNfT799FOcO3cO69atw+nTp+Hj4wN7e3s8f/4cd+7cwfnz57Fx40bpFpGC+uOPP3DgwAEIIRAbG4vAwECcPHkSsbGxqFmzJjZv3qwxoWvXrl3h7OyMH3/8ETdu3ECtWrVw9+5d7NmzBz169Mjy9sjWrVtjy5YteP/991G3bl1oa2ujW7du8PDwwJgxYzBmzBjUrVsXH3zwAVJTU3H48GEIIVCnTh2NW5SePn2KunXrok6dOvDw8EDFihURHh6OXbt2ISUlBZMmTZLqdujQAd9++y2+++47uLq6okOHDnByckJ4eDju378Pf39/fP/993B3dweQ++uNiIjeEVnWYiGiIpG+3GX79u2zrbNs2TIBQAwcOFAqS01NFb///rto2rSpMDExEbq6usLR0VF06NBBrFixQsTFxUl1s1qiNV1WSz2my2oJzZzaEkKICxcuiGHDhglXV1ehr68vdHV1hbOzs+jfv784fPiwRt3Y2FgxceJE4ezsLHR1dYWbm5v47rvvRHJycqYlAm/duiW++OILUbduXWFpaSl0dXWFi4uLGDx4sLh582aeY8xqWcecABDVqlXLU10hhFizZo1o0KCBMDAwEAYGBqJBgwZZLjubUX7OmVzS0tJEpUqVhEKhEA8fPsyx7uHDhzWWg01/jQ8ePDjL+m8/10K8WQazd+/ewtraWjqPW7ZskZYznTFjhkb9vC73euDAATFo0CBRrVo1YWxsLIyMjESNGjXE1KlTxcuXLzX2f/u9kZycLObPny/atWsnKlWqJFQqlbCxsREtWrQQGzduzHUJyoyxIoflXgGIHTt2CCGEGDVqlAAgvv/++0zthIaGigoVKghzc3NpadOM58ff31+0bNlSGBoaChMTE9GjRw8RGBiYqZ3s/gakpKSIBQsWiNq1awtdXV1hbGwsWrZsmeWSqXlZLnPr1q2iXr16wsDAQFSoUEH06dNHPHjwINv36969e0WDBg2Evr6+dF4yCg4OFkOHDhV2dnZCR0dH2NnZiaFDh4rg4OA8H2O6nP4u7Nq1S7Rs2VIYGxsLXV1dUbt2bbFw4UKRkpKSbXtZOXXqlGjRooXQ19cXFhYWonfv3uLRo0c5xvb3338LHx8fYW5uLpRKpahYsaLw9vYWCxcu1Hi9FnS51/QfbW1tYWZmJmrUqCEGDBgg/vnnH43ldzN6+PCh6NWrl7CyspLem5s3b872vfns2TPRp08fUaFCBaGlpaURp1qtFr/99puoWbOm0NPTE7a2tmL48OHixYsXmc5LZGSkmDlzpmjRooWws7MTKpVK2Nvbiw4dOoj9+/dnGevhw4dF165dhZWVlVAqlcLW1lZ4eXmJ7777TmM5YCFyf70REVHxUwjxjmaNIyIiohLrxIkTaNWqFWbMmFFibmEiIiIiygvOsUFEREREREREpRYTG0RERERERERUajGxQURERERERESlFufYICIiIiIiIqJSiyM2iIiIiIiIiKjUYmKDiIiIiIiIiEotJjaIiIiIiIiIqNRiYoOIiIiIiIiISi0mNoiIiIiIiIio1GJig4iIiIiIiIhKLSY2iIiIiIiIiKjUYmKDiIiIiIiIiEotJjaIiIiIiIiIqNRiYoOIiIiIiIiISi0mNoiIiIiIiIio1GJig4iIiIiIiIhKLSY2iIiIiIiIiKjUYmKDiIiIiIiIiEotJjaIiIiIiIiIqNRiYoOIiIiIiIiISi0mNoiIiIiIiIio1GJig4iIiIiIiIhKLSY2iIiIiIiIiKjUYmKDiIiIiIiIiEotJjaIiIiIiIiIqNRiYoOIiIiIiIiISi0mNoiIiIiIiIio1GJig4iIiIiIiIhKLSY2iIiIiIiIiKjUYmKDiIiIiIiIiEotJjaIiIiIiIiIqNRiYoOIiIiIiIiISi0mNoiIiIiIiIio1GJig4iIiIiIiIhKLSY2iIiIiIiIiKjUYmKDiIiIiIiIiEotJjaIiIiIiIiIqNRiYoOIiIiIiIiISi0mNoiIiIiIiIio1GJig4iIiIiIiIhKLR25A6CSIy0tDYsXL0ZcXBzatGmD5s2byx0SERERUZnE6y4ioqJTqkZszJw5EwqFAgqFAkOGDJEtjhMnTkhxODs7S+XBwcFSuUKhkC2+jJydnaV4Tpw4kWPdr776Cl9++SXu3r2Lxo0bv5sAZZKf85ITb29vqR1fX9932ndRyvi6DQ4OzrX+u3ytl8T3FRFRecDrrvwridddJeV5pNKnJL7HiLLzzkds+Pr6YujQodJjhUIBlUoFU1NTVKpUCfXr18eQIUPg5eVVLP0HBARg586dAN7858M/8G+sX78eixYtwsiRI/HLL79AS6tU5byoFON7koio+PC6q2TidRcRUdGS/VYUIQSSkpLw4sULvHjxApcvX8bKlSvRv39/rFy5EoaGhlLdYcOGwcfHBwBgY2NToP4CAgIwa9YsAEDLli0L9B9s3bp14e/vDwDQ09MrUBzvytatW5GYmAgAqF27dpZ1UlJSEBoailWrVuHjjz9+l+HJJi/nhbJnZ2cnvQcKK7f3ZFH2RURU3vG6q3iVxOuuongeiYhKOtkTG/7+/khJSUFwcDD+/vtvHDx4EACwceNGREVFYc+ePdLQJ0dHRzg6OsoWa0pKCoQQMDU1RbNmzWSLIz/q16+fax2lUomvv/76HUQjv7i4OBgZGeXpvFD2dHV139l74F32RURU1vG6q3iVxOsuuZ/HnMTHx2sk06ho8LxSeST7uLdmzZqhVatWGDp0KA4cOIDvvvtO2rZv3z5s2bJFepzdPYJJSUmYPXs26tSpA0NDQ6hUKtja2sLLywtjx45FWFgYgDfDLzMOxzx58mSm+8bevpfs2bNnGDJkCKytraGrq4tbt25le6/n26KjozFmzBjY2dlBT08P9evXh5+fn0YdX19fqS1vb2+NbUOGDJG2zZw5U2NbUlISli5dimbNmsHc3BwqlQr29vbo0qULzp49K9XL6V7Phw8fYuTIkXB1dYWenh6MjIxQp04dTJ8+HVFRURp13z73p0+fRuvWrWFoaAhTU1P07dsXL168yPZcpJsyZYrUzqeffpppu6urq7T90KFDAIBNmzahe/fucHV1hZmZGZRKJSwtLdGyZUusWbMGQgiNNjLOe7F27Vr8/PPPcHd3h0qlwrRp03I8L/nt623r169HnTp1oKenh0qVKmHy5MnSNzd58eTJE4wbNw7Vq1eHvr4+jIyM4OnpicWLFyMlJSXP7Qgh4Ovri1atWsHCwgJKpRK2trbo3r07jh49muO+SUlJmD59OpydnaGnp4eaNWti9erVGnVyu+fyv//+Q79+/eDg4ACVSgVzc3P4+Phkev0X5D0JAFFRUdDX14dCoYC2tjaePn2q0e6GDRuk+k2aNCmS80JEVBbwuqt8XXdl1VZ25+LmzZvo3r07TE1NYWhoiE6dOuH+/fuZ2jt16hSaNm0KfX192NjYYOTIkYiMjMx2vq6M5+TQoUOYMWMGXFxcoKOjg1WrVkn18nrtAAB37tzBgAEDpLqGhoZwdnZGly5dsHTp0kz1/fz80KlTJ1hbW0OpVKJChQpo164dtm7dmqnuuXPn0L17d9jZ2UGpVMLExASurq7o1asX/vrrrzyd84zn4tatW5gyZQocHR2hq6ub5XVVQeLM63nNTlhYGCZPngwPDw8YGRlBX18fLi4uGDhwYJ5eWzNnzkSbNm3g6OgIIyMj6T3Ro0ePLOeX2b9/P9q1awcrKysolUqYmZmhevXq6N+/P/bv369RNy0tDb///juaN28uvd+cnJwwYsQIBAUF5RoblVPiHVu7dq0AIP28LS0tTbi5uUnbO3bsKG2bMWOGVD548GCpfNCgQRptvv1z9uxZIYTIsU56LEFBQRplGWMBIK5cuSKOHz8uPXZycpLieHvfunXrZupDoVCIv/76K8vz0bJlS41zMXjwYGnbjBkzpPLw8PAs207/Wbx4sVTXyclJKj9+/LhUfuLECWFkZJRtG5UrVxZPnjzJ8txXrlxZ6OjoZNqnffv2uT7/9+7dk+pbWFiI5ORkadvZs2c1zmtaWpoQQoi+ffvm+Lx98cUXGn20bNky2+cvvW5256UwfWX3nHTo0EGo1epcn5OzZ88KMzOzbPtu1aqVSExMzPUcp6amivfffz/H45gzZ47GPrm9bgGIuXPnSvXffq1ntHz5cqGlpZVt31OmTMmy37y+J9MNGDBAKluwYIFGDO3bt5e2rV69usDnhYiotON1V/m+7srpecx4Luzs7IShoWGmPmrUqCFdjwkhxJEjR4RSqcxUr169ehqPg4KCsjwnbz+/6ecuP9cOr169EhYWFtnWrVatmsbxjxkzJsfX4SeffCLVvX37ttDV1c22bl7P+dvnMKu2Ml5X5TfOvJ7X7Fy4cEFYWlpm29eVK1eEEDlf79nY2GS7v0KhENu2bZPqHj16VCgUimzrf/rpp1LdhIQE0apVq2zrmpmZifPnz+fpeaDyRfYRG2/T0tJCmzZtpMcXL17MdZ9t27YBAExNTbF27VocO3YMmzdvxsyZM9GgQQNpQiZ/f39MnTpV2u+9996Dv7+/9JOVkJAQzJ49GwcPHsTKlStRoUKFPB/L8+fP4evrix07dqBRo0YAACEERo0ahfj4+Dy387bRo0fjypUrAACVSoUvv/wSe/fuxebNmzF8+HDo6urmuH9iYiL69++PuLg4AEDDhg2xfft2rF+/HhUrVgQABAUF4ZNPPsly/6CgILRq1Qp+fn6YMWOGVH7w4EHcvXs3x77d3NzQokULAEBERIRGhnbjxo3S70OHDpWet27duuG3336Dn58fjh8/jqNHj2L16tXSc/HLL79I3w69LTAwEN26dcOOHTuwc+dOtG3bNsf4CtNXQEAAvvrqK+zbtw8TJkyQyg8cOKBxbFlJSkpC3759pW9sevXqhb1792Lr1q3w8PAAABw/fhxz5szJsR0AWL58uTRRm1KpxOzZs7Fv3z6N+3i/+eYb/Pfff1nu//DhQ/zyyy/YvXs3OnfuLJVPnz4djx49yrHvmzdvYsyYMVCr1dDS0sI333yDQ4cO4ffff4e5uTkAYN68eTh27BiAgr0n0w0fPlz6PeM3KC9evJBGXxgZGaFPnz5Fcl6IiMoiXnflrjRfd+XVs2fPUK1aNWzbtg0///wzlEolAODWrVs4fPgwAECtVuOzzz6TRpBWr14d//zzDzZs2IDnz5/nqZ/AwEAMHToUe/bswZYtW+Dp6Znva4fjx48jIiICANCqVSvs2bMHBw8exJo1a/Dxxx9L5xR4MwJi2bJl0uPx48dj3759mDx5sjRqaOXKlfjnn38AAHv27EFSUhIAoHfv3jhw4AD27duH33//HQMGDMjX6zHd06dPc72uym+ceTmv2UlKSkLv3r0RHh4OALC2tsbixYulc9i2bds8rYAybtw4rFu3Dnv37sWJEydw8OBBzJ07F8Cb9923334r1d2+fbs06vnzzz/HkSNH4Ofnh19++QU9evSAiYmJVHfmzJk4fvw4AKBy5cpYu3YtDh06hM8++wzAm1G7/fr1Q2pqaq4xUjnzrjMpuX1zIIQQ33zzjbRdqVRK5dllnO3t7QUAYW9vL06fPi3i4uLy1P/bmXohMmcmly5dmqlOXr852Lt3r7QtNDRUqFQqadv27dtzjSerbw6ioqI0svZLlizJ9liFyPqbg127dkllKpVKhIaGSvX37NmjkW19/vy5EELz3FeoUEEkJCRI+1SvXl3a5ufnl2M8Qgixbt06qX6fPn2EEG++Tbe2thYAhJaWlnj06JFU/9WrV2Ly5Mmidu3awtDQMMuMb8Z+M46i8PT0zPN5KWxfvXv31uijS5cu0rZu3brl2Pfu3bulMisrK3Hq1Cnh7+8v/P39xbJly6RtdnZ2uZ7fjN8qjRkzRmNb/fr1pW2ff/65VJ7x+JYvXy6VJyQkCCsrK2nbokWLhBDZZ/AnTpwolfn4+EjH4O/vL4YNGyZt+/DDD6V98vueTKdWq4WLi4tUfuvWLSGEEEuXLpXKhg8fXqjzQkRU2vG6i9ddeRmxoVQqNUaMdOjQIdNzcvHiRY3zffny5SyPA8h+xEbPnj0zxZffa4dDhw5JZf379xeBgYEiNTU1y2Pv0aOHVLdr164a2z744ANpW6dOnYQQQqxcuVIqmzhxonj06JHGqNu8yu91VX7jzMt5zU7G50pLS0vjeXxbTiM2bt68KQYNGiQqV66c7SiXmJgYIYQQU6dOlcoWLlyo8R7ISK1WZzo/GV8PdnZ20rYDBw7k+ZipfChxIzYA4OXLl9LvZmZmudZPz+CFhoaiadOmMDIyQqVKldC9e3f8/fffhYqlV69eBd4340RXdnZ2cHFxkR4HBgYWqM179+5pZCh79uyZ7zbu3Lkj/V6lShXY2dlJjzPGLITI8psALy8v6OvrS48tLS2l39Mz6Dn54IMPpMzs7t27ERsbiyNHjkj38/n4+EiTXL1+/RpNmzbF/Pnzcf36dcTHx2c5z0VkZGSWfeXn/BS2r7cnNsv4OLfn+9atW9LvL1++RIsWLdC8eXM0b94cY8aMkbY9e/ZMyrBnJ+Pzm1NMGetlV0dfX18j65+f4zhy5Ih0DM2bN8eaNWukbTdu3MixnbxQKBQYNmyY9Dh91EbG0TEZR3UU9rwQEZVVvO7KXlm47sqL6tWra4x0yKqPjOfQwMAAdevWlR7ndXLXrJ7f/F47NG/eHDVr1gTw5v98Nzc3GBgYoFatWvj888812svv//3du3eHra0tAGDhwoVwcnKCoaEh6tWrh0mTJuHx48d5Os7s+snuuqqw1yj5ed9kPD+VK1fWeB7z6vr162jYsCHWr1+PoKAgaZTL29KvmQcOHChNZjpx4kTY29vD2NgYXl5emDlzpvQae/nypcbfowkTJmi8Hp49eyZtK4prSSpbSlxiQ61W48iRI9LjvMwu/e2338LPzw+DBw9G3bp1YWRkhKdPn8LPzw8ffvghlixZUuB4Mv7nUxwyDvV6e0hVxjd2SWJhYaHxWEfnf4vrZJUIeJuBgQE+/PBDAG+SCdu3b8/2w+iOHTuk/+QNDQ2xdOlSHD9+HP7+/hrLqKnV6iz7ys/zV9i+3pX0oaylWVEdw+DBg6Uhzxs3bsSDBw9w7tw5AECNGjXg5eVVJP0QEZVVvO76n7J63VVUfeTl9oTcFOb5Tb920NPTw+nTp7FgwQJ07twZVapUQVpaGm7evIkVK1agSZMmCAkJKVAf1tbWuHz5MmbPno22bdvC0dERiYmJuHLlChYuXIjmzZsjJiamwMdQXIr7ffO2ZcuWSbd3ubm54a+//sKpU6cyTRqafs1cvXp1BAQEYMqUKWjZsiXs7OwQFxeHc+fOYdasWWjfvj3S0tLyFUNZuB6molXiEhvfffcdHj58KD0ePHhwrvsIIdC1a1f4+vri8uXLiImJ0ZjVe9OmTdLv6R+CgLx9QC3MH/HTp09Lv4eFhWkcl6urKwBI9w4C0FjZISoqCv/++2+mNqtWrQptbW3p8Y4dOzLVye0/uerVq0u/P3jwQGPOiIwxKxQKVKtWLce2Cipj8uKPP/6QjsPS0hLvv/++tC3jf0wdOnTAmDFj4O3tDQ8PDzx58iTXfvLz/BW2r4zn7u3H6c93dtzd3aXfHR0dpSXu3v6Ji4uDk5NTjm1lfH5ziiljvezqJCYm4vLlywU6jn79+mV5DEIIjSx7ft+TGVWqVAnt27cH8OYe5LFjx0rbMr7GgMKfFyKisojXXW+U9euuouDm5ib9npCQgOvXr0uPc5sXK11Wz29+rx3E/y8BPHHiROzZswf3799HTEyMNGohOjoa+/btA5D///uFELCzs8O3336LQ4cO4dGjR4iIiJC+KHn06BHOnDmTp2PNqp/srqsKe42Sn/dNjRo1pN+DgoJw9erVTHVye01nvGYeO3Ys+vfvj+bNm2u8V95uz9XVFXPnzsWJEycQGhqKZ8+eSascXbx4EYGBgbCystKYx+TgwYPZXg9nnG+GCAB0cq9SvP7991+kpKTg0aNH2LRpk7TEJwB06tRJmvgvJ82aNUP16tXh5eUFe3t7KJVKHDhwQNqecbnNjEPrrl27hu3bt8Pa2hpmZmaoVatWER3VG5988gnmzJkDU1NTzJ8/H8nJyQDeDPNs164dgDf/YaYLDg7GkCFDUL9+faxevTrLjLCpqSl69+6NzZs3AwC+/PJLPH36FC1btkRcXByOHj2KOnXqYOTIkdnG1a5dO9jb2yM0NBTJycno0aMHJk+ejLi4OEyZMkWq17FjR1hbWxfJuXhbw4YNUatWLdy4cUPjQmLgwIFQqVTS44zDSI8ePYoNGzbA1NQUCxYsyPaWkIIqbF9bt26VMtHHjh3Dnj17pG25vY7btm0LBwcHPH78GCEhIWjfvj1GjBgBa2trPHv2DA8ePMChQ4fg5uaGtWvX5tjWkCFDpEnOfvvtN1hbW8PT0xM7duzAhQsXpHrZXbymT/Tm6OiI33//XbpFSFtbO9chuEOGDMHPP/8MtVqNTZs2wdjYGF26dIGuri6ePHmCW7duwc/PD1OnTpWWnCvse3L48OHSJLTpFzJKpRIDBw4s0vNCRFQW8LqrfF53FYW6devCzc1Nun2if//+mDlzJhITEzF58uQCt5vfa4f//vsPI0aMQI8ePVCtWjXY2toiIiICN2/elNpMfw0OGTJESkbt3r0bkyZNgo+PD06dOiVNgpteDwD++ecfLFq0CN27d4eLiwusra0RGhqqscRoxtd3XuTluiq/cRaGj48PnJyc8OjRI6jVanTo0AFTpkyBu7s7nj17hs2bN2PevHmoU6dOtm1kvGb+448/4OzsjIiICEybNi3L+gsXLsT+/fvRuXNnODk5wcLCAoGBgRqjpBITE6Ulon/66ScAwKBBg/D111+jVq1aiIuLQ0hICM6dO4c9e/aUyJEzJLNincEjC29PYpXdz4ABAzJNRpXd5EfVqlXLsa30iXmEECIiIkIYGBhkqtOmTRshRM6T5KTL6yRWGSd3Sv9RKBRi/fr1Gu21bds2Uz2VSiWqVq0qPc647NirV6+Eh4dHtsdbVMuOPX78ONdzL4TmBJpr167N8pxlZfHixZn6vX79ukad+Ph4jQki039sbW01zm/GfvMST1bnpbB9ZfV8AxBt27bVWCotu+fkzJkzOS73mtW5z0peljX9/vvvNfbJ7XULQHz33XdS/ZzeJ7/88kuOS7a9fQ4L+55MTk7WmGgKgOjVq1eRnBciotKO11287srL5KF5Xfo2u+Ve33vvPY3H2U0emvGcZJSfa4ezZ8/mWM/Y2FgEBwdLbY8ePTrH+iNGjJDqbtq0Kce6lSpVkibEzEnGfbJbKjjjdVV+48zrec3O+fPnhbm5ebZ95bbc67Vr17J8HXh7e2f5Opg3b16Ox1a3bl3pWjkhISFTO1n9EL1N9ltRFAoFVCoVrK2tUa9ePXzyySc4c+YM/vzzT2mSmdx8/fXX6N27N1xdXWFiYgJtbW1YWFjA29sbGzZswPjx46W65ubm2L59O+rXr5/r8lyF5e/vj48//hjW1tbQ1dVFvXr1sG3btkzfJK9fvx59+vSBiYkJDAwM0KZNG5w6dSrb+QEsLS1x/vx5LFq0CF5eXjA1NYVSqYSdnR06deokLXGWk5YtWyIgIACffvopXFxcoFKpoK+vj9q1a2PatGm4fPkyKlWqVCTnITsfffSRxuiM9FEcGRkYGODYsWPo0aMHLCwsYGpqim7duuHff/+FjY1NkcZT2L4mT56MFStWoEaNGlCpVLC3t8ekSZOwa9cujaG42fHy8sL169cxYcIE1KxZEwYGBtDX10flypXRtm1bLF68GLNnz861HW1tbWzfvh1r1qxBy5YtYWZmBh0dHVhbW6Nbt244cuQIvvnmm2z337p1KyZOnIhKlSpBpVLB3d0dv//+e7ZZ+LeNGjUK586dw4ABA+Do6AiVSgUTExNUq1YNvXv3xvr16zVGfhT2PZnV6Iy3b0MBCn9eiIjKAl53ld/rrqLQpk0bHD58GE2aNIGuri6srKzw8ccfa4wqAJDn11K6/Fw7uLq64ptvvkHLli1hb28PXV1dKJVKODo64qOPPsL58+c1bttdtmwZduzYgQ4dOqBChQrQ0dGBhYUFfHx8sGXLFqxcuVKq26hRI0yaNAleXl6wtbWFSqWCrq4uXF1d8dlnn+Hs2bMwNjbO17Ft2rQJ06ZNk44ru+uq/MRZWA0bNsSNGzcwadIk1KpVCwYGBtDT00PlypUxYMAA2Nvb57h/7dq1cfjwYXh5ecHQ0BC2trYYPXo0du/enWX9Dh06YPTo0ahXrx6sra2ho6MDfX191KhRA19++SWOHj0qXSvr6+vjyJEjWLlyJby9vWFhYQEdHR3Y2NjA09MT48ePzzSXBxEAKIQoolmHiKjcuH79Ojw8PAC8mWgrr+vXExERUeklhMhyPofdu3ejW7duAN5MRPry5cs8falTVmU8R0FBQdJcEkRUfMrvXxwiyre0tDS8evUKq1evlsoyTkJFREREZde9e/fQoUMHbNu2DXfv3sWDBw+wceNGaQlg4M2I3PKc1CAieXDEBhHl2YkTJ9CqVSuNsv3796NDhw4yRURERETvyp07dzRWMXlbo0aNcPjw4XzfrlHWcMQG0bvHdCoR5ZuOjg7c3d3h6+vLpAYREVE5YW1tjREjRqBmzZowNTWFjo4OLC0t4e3tjRUrVsDf37/cJzWISB4csUFEREREREREpRZHbBARERERERFRqcXEBhERERERERGVWkxsEBEREREREVGppSN3AERERPTuqNVqhIaGwtjYWGPmfiIiIqKSRAiB2NhY2Nvb57qMNBMbRERE5UhoaCgcHBzkDoOIiIgoTx4/foxKlSrlWIeJDSIionIkfSnGx48fw8TEROZoiIiIiLIWExMDBweHPC0jzcQGERFROZJ++4mJiQkTG0RERFTi5eXWWU4eSkRERERERESlFhMbRERERERERFRqMbFBRERERERERKUW59ggIiIiIiKiciMtLQ0pKSlyh0EAVCpVrku55gUTG0RERERERFTmCSEQFhaGqKgouUOh/6elpYXKlStDpVIVqh0mNoiIiIiIiKjMS09qWFtbw8DAIE+rbVDxUavVCA0NxbNnz+Do6Fio54OJDSIiIiIiIirT0tLSpKSGpaWl3OHQ/7OyskJoaChSU1OhVCoL3A4nDyUiIiIiIqIyLX1ODQMDA5kjoYzSb0FJS0srVDtMbBAREREREVG5wNtPSpaiej6Y2CAiIiIiIiKiUotzbBARERERUak1e/+lQu0/vaNnEUVCVDyGDBmCdevWAQCUSiUcHR0xaNAgTJ06FTo6/EgPMLFBRERERERE5Vhhk2P5UdBEWocOHbB27VokJSVh3759GDVqFJRKJaZMmVLEEZZOvBWFiIiIiIiIqATT1dWFra0tnJycMHLkSPj4+MDPzw+LFi1C7dq1YWhoCAcHB3z++eeIi4vT2HfVqlVwcHCAgYEBevTogUWLFsHMzEza/uDBA3Tv3h02NjYwMjJCgwYNcOTIkXd8hIXDxAYRERERERFRKaKvr4/k5GRoaWlh6dKluHnzJtatW4djx47hq6++kuqdPn0an332Gb744gsEBASgbdu2mDNnjkZbcXFx6NSpE44ePYorV66gQ4cO6Nq1K0JCQt71YRUYExtEREREREREpYAQAkeOHMHBgwfRunVrjBs3Dq1atYKzszNat26N77//Hlu2bJHqL1u2DB07dsSkSZNQtWpVfP755+jYsaNGm3Xq1MGnn36KWrVqwc3NDd999x2qVKkCPz+/d314BcbEBhEREREREVEJtmfPHhgZGUFPTw8dO3ZE3759MXPmTBw5cgRt2rRBxYoVYWxsjIEDByI8PBwJCQkAgLt376Jhw4Yabb39OC4uDpMmTYK7uzvMzMxgZGSE27dvc8QGERERERERERWNVq1aISAgAIGBgXj9+jXWrVuHly9fokuXLvDw8MC2bdtw6dIlLF++HACQnJyc57YnTZqEHTt2YO7cufD390dAQABq166drzbkxlVRiIiIiIiIiEowQ0NDuLq6apRdunQJarUaCxcuhJbWmzELGW9DAYBq1arhwoULGmVvPz59+jSGDBmCHj16AHgzgiM4OLiIj6B4ccQGERERERERUSnj6uqKlJQULFu2DA8fPsSGDRvw22+/adQZM2YM9u3bh0WLFiEwMBC///479u/fD4VCIdVxc3PD9u3bERAQgKtXr6J///5Qq9Xv+nAKhYkNIiIiIiIiolKmTp06WLRoEebPn49atWrhr7/+wrx58zTqNG3aFL/99hsWLVqEOnXq4MCBAxg/fjz09PSkOosWLYK5uTmaNGmCrl27on379qhXr967PpxCUQghhNxBEBER0bsRExMDU1NTREdHw8TERO5wiIgKbfb+S4Xaf3pHzyKKhEqyxMREBAUFoXLlyhof6sujESNG4M6dO/D395c7lByfl/xcs3CODSIiIiIiIqIyasGCBWjbti0MDQ2xf/9+rFu3Dr/++qvcYRUpJjaIiIiIiIiIyqj//vsPP/74I2JjY+Hi4oKlS5fi448/ljusIsXEBpUKyalpiE5MRkxiCmITk5GSpoZaAGoh4B4bAf20FCi0tABtHSgM9KBlaACFoSEUhgYaE+MQERERERGVJ2+vlFIWMbFBsguPT8STyHg8joqT/n0WnYCo18mISUxGbGIKElPTst1/z9MLSHoQlPVGLQUU+vpQGBpAy9QU2tYVoG1jBW0ba2jbWEHL2grattbQ0i/f99kRERERERGVVkxs0DuTmJKGO88jcfNZJG6FRSLwRTQeR8UjITm1+DpVC4j4BIj4BKhfvEJq4IMsq2lZV4DSxRk6//+jrOIMbVvr4ouLiCgLT58+xeTJk7F//34kJCTA1dUVa9euRf369QEAQgjMmDEDq1atQlRUFJo2bYoVK1bAzc1N5siJiIiI5MPEBhWbh69icOnxK9x8FoGbzyLx8FUMUtUlcxEe9YtXSHrxCknnLkplCkMDKN2qQFnbHSqPmlBWc4NCybcMERWPyMhING3aFK1atcL+/fthZWWFwMBAmJubS3V+/PFHLF26FOvWrUPlypXx7bffon379rh161a5n+GdiIiIyi9+SqMiE5OYjPPBL3Dm4XOcDXqOZzEJcodUKCI+AckB15EccB3xAKCrC5V7VSg9akBVpxaU1d3ezOtBRFQE5s+fDwcHB6xdu1Yqq1y5svS7EAI///wzpk2bhu7duwMA1q9fDxsbG+zcuRMffvjhO4+ZiIiIqCRgYoMK5d6LKBy7G4ozQWG4ERqJNFEyR2QUiaSkDImOv6FlZgrdxp7Q9WoIVV0PjuYgokLx8/ND+/bt0bt3b5w8eRIVK1bE559/jhEjRgAAgoKCEBYWBh8fH2kfU1NTNGrUCGfPns02sZGUlISkpCTpcUxMTPEeCBEREdE7xk9ilG9PouKx/2YI9t96jAevyu8FsjoqGq8PHMPrA8egMNCHqkFd6DVtBN2GnlDoquQOj4hKmYcPH2LFihWYMGECpk6digsXLmDs2LFQqVQYPHgwwsLCAAA2NjYa+9nY2EjbsjJv3jzMmjWrWGMnIiIikhMTG5Qn4fGJOHDrMfbfeozroRFyh1PiiITXSDp5Bkknz0BhaAC9lk2h3741lFWryB0aEZUSarUa9evXx9y5cwEAdevWxY0bN/Dbb79h8ODBBW53ypQpmDBhgvQ4JiYGDg4OhY6XiIiIqKRgYoNydCnkJTZevI/j90LL9m0mRUjEJ+D1vsN4ve8wdCo7Qb99K+i1bgEtYyO5QyOiEszOzg41atTQKHN3d8e2bdsAALa2tgCA58+fw87OTqrz/PlzvPfee9m2q6urC11d3aIPmIiIiN6JIUOGICoqCjt37pQ7lBKLiQ3KJDElDftuhWDzxfu4+yJa7nBKtdSgR4j9zRexq/+CXssmMPygG3Sc+E0pEWXWtGlT3L17V6Ps3r17cHJyAvBmIlFbW1scPXpUSmTExMTg/PnzGDly5LsOl4iIqMyIWbrynfVlMvaTd9ZXecIlHUjyPCYBPx+/jvbL92LWvktMahSllBQkHjmJ8JGTEDl9HpKv3ZQ7IiIqYcaPH49z585h7ty5uH//PjZu3IiVK1di1KhRAACFQoFx48bh+++/h5+fH65fv45BgwbB3t4e77//vrzBExER0Tvh7e2NMWPGYNy4cTA3N4eNjQ1WrVqF+Ph4DB06FMbGxnB1dcX+/fulfU6cOAGFQoG9e/fCw8MDenp6aNy4MW7cuCHVCQ8PR79+/VCxYkUYGBigdu3a2LRpk0bfsbGxGDBgAAwNDWFnZ4fFixfD29sb48aNk+pERkZi0KBBMDc3h4GBATp27IjAwMBiPy9MbBCex77GnAOX0fm3A1h77i6iXifLHVLZJQSSL1xB5ORZCP9iChL9z0Ko1XJHRUQlQIMGDbBjxw5s2rQJtWrVwnfffYeff/4ZAwYMkOp89dVXGDNmDD755BM0aNAAcXFxOHDgAPT09GSMnIiIiN6ldevWoUKFCvjvv/8wZswYjBw5Er1790aTJk1w+fJltGvXDgMHDkRCQoLGfl9++SUWLlyICxcuwMrKCl27dkVKSgoAIDExEZ6enti7dy9u3LiBTz75BAMHDsR///0n7T9hwgScPn0afn5+OHz4MPz9/XH58mWNPoYMGYKLFy/Cz88PZ8+ehRACnTp1kvopLgohOHFCefUqLhGrz97BtoCHSEotvR+u9zy9ANWDILnDKDBtx0owGtQXek0byR0KEZUDMTExMDU1RXR0NExMTOQOh4io0Gbvv1So/ad39CyiSKgkS0xMRFBQECpXrpzpC4GSfitKxjk2vL29kZaWBn9/fwBAWloaTE1N0bNnT6xfvx4AEBYWBjs7O5w9exaNGzfGiRMn0KpVK2zevBl9+/YFAERERKBSpUrw9fVFnz59suy3S5cuqF69OhYsWIDY2FhYWlpi48aN+OCDDwAA0dHRsLe3x4gRI/Dzzz8jMDAQVatWxenTp9GkSRMAb0aCODg4YN26dejdu3emPnJ8XvJxzcI5NsqhyIQkrD13F39ffoDElDS5wyn30kKeIPr7hYh3qwKjof2hW7e23CEREREREVEJ5eHhIf2ura0NS0tL1K79v88Q6UvDv3jxQmM/Ly8v6XcLCwtUq1YNt2/fBvAmQTJ37lxs2bIFT58+RXJyMpKSkmBgYADgzbL0KSkpaNiwodSGqakpqlWrJj2+ffs2dHR00KjR/76wtbS01OinuDCxUY4kp6Zh/X/3sPrsXSQkp8odDr0lNfABoqZ+B1X992A07CMoKzvKHRIRERGVAxzxQFS6KJVKjccKhUKjTKFQAHizlHxe/fTTT1iyZAl+/vln1K5dG4aGhhg3bhySk0vHNAWcY6OcOH4vFD1WHcKykzeZ1Cjhki8GIGL0l4j5dTXU8Qm570BERERERJSLc+fOSb9HRkbi3r17cHd3BwCcPn0a3bt3x0cffYQ6derAxcUF9+7dk+q7uLhAqVTiwoULUll0dLRGHXd3d6SmpuL8+fNSWXh4OO7evZtpSfuixhEbZdyTqHj8cOgK/B+EyR0K5Yda4PXug0g6/R+MRwyCnndTuSMiIiIiIqJSbPbs2bC0tISNjQ2++eYbVKhQQVpZzc3NDVu3bsWZM2dgbm6ORYsW4fnz51JCwtjYGIMHD8aXX34JCwsLWFtbY8aMGdDS0pJGiLi5uaF79+4YMWIEfv/9dxgbG+Prr79GxYoV0b1792I9No7YKKNS0tRYefo2eq06xKRGKaaOiET0/CWInDYHqaF8HomIiIiIqGB++OEHfPHFF/D09ERYWBh2794NlUoFAJg2bRrq1auH9u3bw9vbG7a2tpmWk1+0aBG8vLzQpUsX+Pj4oGnTpnB3d9eY9HPt2rXw9PREly5d4OXlBSEE9u3bl+n2maLGERtlUOCLaEzd/R/uvYiWOxQqIsmXriJ85EQY9esFg97vQ6HNnCQRERERUVEoyEol75Kvr6/0+4kTJzJtDw4OzlSW1eKnzZo1w40bN7Lsw8LCAjt37swxDmNjY/z111/S4/j4eMyaNQuffPK/82dubi6tzvIuMbFRhqiFwIb/7uGXkzeRnFZ6l2+lbCSnIG7dZiRduAKTSaOhY2cjd0RERERERFROXLlyBXfu3EHDhg0RHR2N2bNnA0Cx32aSF0xslBHPohMwbc8FXAx5KXcoVMxSbt1FxKivYDxyKPTbessdDhERERERlRMLFizA3bt3oVKp4OnpCX9/f1SoUEHusJjYKAt2X3+EHw5fQVwSVzspL8Tr14hZ9CuSLlyGyZhPoGVsJHdIRERERERUAnl7e2d5a0p+1a1bF5cuFW556OLCxEYplpSahu/2X8buG4/kDoVkkuR/DuG37sF06nioalSTOxwiIiIiIqJ3jjMQllJhMQkYsuEEkxoEdXgEIifPQsK+w3KHQkRERERE9M4xsVEKXQx5iQ/XHsWtsEi5Q6GSIjUVsctWIWbpSogU3pJERERERJSVorglg4pOUT0fTGyUMn9dCMSnm04hMiFJ7lCoBHq9/wgiv56FtIgouUMhIiIiIioxlEolACAhIUHmSCij5ORkAIC2tnah2uEcG6VESpoas/dfgt913npCOUu5dRcRYyfDbOZkKF1d5A6HiIiIiEh22traMDMzw4sXLwAABgYGUCgUMkdVvqnVarx8+RIGBgbQ0SlcaoKJjVIgITkVE7afxdmg53KHQqWEOjwSkV/Ngum3k6Bbt7bc4RARERERyc7W1hYApOQGyU9LSwuOjo6FTjIxsVHCRSUkYfQ/p3E9NELuUKiUEa9fI2r6PJh+ORp6LZrIHQ4RERERkawUCgXs7OxgbW2NlJQUucMhACqVClpahZ8hg4mNEux5TAI+2+yPh+GxcodCpVVqKqLnL4E6OgYGXTvIHQ0RERERkey0tbULPacDlSycPLSECg6PxeANJ5jUoMJTC8T+ugZx6/+WOxIiIiIiIqIixxEbJdCd51H4bNMpRL5OljsUKkPiN22DSEmB8fCP5A6FiIiIiIioyHDERgnz8FUMPtvsz6QGFYuErX4cuUFERERERGUKExslyJPIOHyy6RQiE5LkDoXKsPhN2xC3aZvcYRARERERERUJJjZKiOcxCRix6RRexiXKHQqVA/Hr/0b8Vj+5wyg31Go1atWqhTlz5hRo/8aNG+Orr74q4qiIiIiIiMoGJjZKgPD4RIzYdAqh0Qlyh0LlSNzqP5Gwa3+Rt+vr6wuFQoGLFy8Wedul1aZNm/D48WOMHj1aKouLi8OMGTPQoUMHWFhYQKFQwNfXN8v9J0+ejOXLlyMsLOwdRUxEREREVHowsSGzmMRkfLrJH48i4uQOhcqh2JW+SDz9n9xhlHk//fQTPvzwQ5iamkplr169wuzZs3H79m3UqVMnx/27d+8OExMT/Prrr8UdKhERERFRqcPEhoxS1Wp8ueMcAl9Gyx0KlVdqgeifliHl7n25Iymzrly5gqtXr6JPnz4a5XZ2dnj27BkePXqEn376Kcc2tLS08MEHH2D9+vUQQhRnuEREREREpQ4TGzL68fBVnAt+IXcYVN4lJSFq1nykPX9ZbF0MGTIERkZGCAkJQZcuXWBkZISKFSti+fLlAIDr16+jdevWMDQ0hJOTEzZu3Kixf/rtLadOncKnn34KS0tLmJiYYNCgQYiMjNSou2vXLnTu3Bn29vbQ1dVFlSpV8N133yEtLS1TXMuXL4eLiwv09fXRsGFD+Pv7w9vbG97e3hr1Xrx4geHDh8PGxgZ6enqoU6cO1q1bl6dj37lzJ1QqFVq0aKFRrqurC1tb2zy1AQBt27bFo0ePEBAQkOd9iIiIiIjKAyY2ZLLl8gP8ffmB3GEQAQDUkdGInPED1PHFN89LWloaOnbsCAcHB/z4449wdnbG6NGj4evriw4dOqB+/fqYP38+jI2NMWjQIAQFBWVqY/To0bh9+zZmzpyJQYMG4a+//sL777+vMYrB19cXRkZGmDBhApYsWQJPT09Mnz4dX3/9tUZbK1aswOjRo1GpUiX8+OOPaN68Od5//308efJEo97r16/h7e2NDRs2YMCAAfjpp59gamqKIUOGYMmSJbke95kzZ1CrVi0olcoCnrk3PD09AQCnT58uVDtERERERGWNjtwBlEfng59j/uEAucMg0pD26DGi5yyE2XffQKFd9DnPxMREfPTRR5gyZQoAoH///rC3t8ewYcOwadMm9O3bF8CbkQnVq1fHunXrMHPmTI02VCoVjh49KiUJnJyc8NVXX2H37t3o1q0bAGDjxo3Q19eX9vnss8/w2Wef4ddff8X3338PXV1dJCcn49tvv0WDBg1w7Ngx6Oi8+VPo4eGBIUOGoFKlStL+K1euxO3bt/Hnn39iwIABUpstW7bEtGnTMGzYMBgbG2d73Hfu3EGjRo0KefaAihUrQqVS4datW4Vui4iIiIioLOGIjXfsUUQsJu04h1Q175Onkif5ynXErd9cbO1//PHH0u9mZmaoVq0aDA0NNeafqFatGszMzPDw4cNM+3/yyScaIx9GjhwJHR0d7Nu3TyrLmNSIjY3Fq1ev0Lx5cyQkJODOnTsAgIsXLyI8PBwjRoyQkhoAMGDAAJibm2v0uW/fPtja2qJfv35SmVKpxNixYxEXF4eTJ0/meMzh4eGZ2iwoc3NzvHr1qkjaopJn5syZUCgUGj/Vq1eXticmJmLUqFGwtLSEkZERevXqhefPn8sYMREREVHJwMTGO5SQnIpxW88gJjFF7lCIspXwzy4k/Xe5yNvV09ODlZWVRpmpqSkqVaoEhUKRqfztuTMAwM3NTeOxkZER7OzsEBwcLJXdvHkTPXr0gKmpKUxMTGBlZYWPPvoIABAd/Wai3kePHgEAXF1dNdrT0dGBs7OzRtmjR4/g5uYGLS3NP5fu7u4abeWkqCb8FEJkOldUttSsWRPPnj2Tfv79919p2/jx47F79278888/OHnyJEJDQ9GzZ08ZoyUiIiIqGXgryjv045EAPAyPlTsMopwJgegFv8Dylx+hbV2hyJrV1tbOV3lBkgFRUVFo2bIlTExMMHv2bFSpUgV6enq4fPkyJk+eDLVane82C8vS0jLLJE1BREVFoUKFontOqOTR0dHJclLZ6OhorF69Ghs3bkTr1q0BAGvXroW7uzvOnTuHxo0bv+tQiYiIiEoMjth4Rw7efowdV4PlDoMoT0RsHKLnLYZITZU7FA2BgYEaj+Pi4vDs2TNplMWJEycQHh4OX19ffPHFF+jSpQt8fHwy3Qri5OQEALh/X3OZ29TUVI3RH+l1AwMDMyVF0m9rSW8rO9WrV89yItT8evr0KZKTk6WRIlQ2BQYGwt7eHi4uLhgwYABCQkIAAJcuXUJKSgp8fHykutWrV4ejoyPOnj0rV7hEREREJQITG+9AaHQ8vttf9EP7iYpTyp1AxK3+U+4wNKxcuRIpKf+7lWvFihVITU1Fx44dAfxv9EfG0R7Jycn49ddfNdqpX78+LC0tsWrVKqRmSN789ddfmUZXdOrUCWFhYfj777+lstTUVCxbtgxGRkZo2bJljjF7eXnhxo0bSEpKyufRarp06RIAoEmTJoVqh0quRo0awdfXFwcOHMCKFSsQFBSE5s2bIzY2FmFhYVCpVDAzM9PYx8bGBmFhYTm2m5SUhJiYGI0fIiIiorKEt6IUszS1wJRd/yE2ifNqUOmTsHMfVPXrQtezjtyhAHiTpGjTpg369OmDu3fv4tdff0WzZs2kFVGaNGkCc3NzDB48GGPHjoVCocCGDRsy3daiUqkwc+ZMjBkzBq1bt0afPn0QHBwMX19fVKlSRWMei08++QS///47hgwZgkuXLsHZ2Rlbt27F6dOn8fPPP+e4IgoAdO/eHd999x1OnjyJdu3aaWz75ZdfEBUVhdDQUADA7t27peVmx4wZA1NTU6nu4cOH4ejoiLp16xb8BFKJlp6gA96s0NOoUSM4OTlhy5YtGpPi5te8efMwa9asogiRiIiIqETiiI1itsL/JgKehssdBlGBxSxeAXVcvNxhAHiTCHB3d8f06dPh6+uLfv36YdeuXVIiwtLSEnv27IGdnR2mTZuGBQsWoG3btvjxxx8ztTV69GgsXboUISEhmDRpEvz9/eHn5wczMzPo6elJ9fT19XHixAkMGDAA69atw8SJExEREYG1a9fiiy++yDVmT09PeHh4YMuWLZm2LViwAN9++y1WrFgBANi+fTu+/fZbfPvttxojR9RqNbZt24ZBgwZx8tByxMzMDFWrVsX9+/dha2uL5ORkREVFadR5/vx5lnNyZDRlyhRER0dLP48fPy7GqImIiIjePYUoqun6KZMboREYuP4YuLJr8drz9AJUDwo/hwFlT69NC5hOGi1b/76+vhg6dCguXLiA+vXrF1s/arUaVlZW6NmzJ1atWlVk7W7YsAGjRo1CSEhIplsJ8mLnzp3o378/Hjx4ADs7uyKLi0q2uLg4ODo6YubMmRg8eDCsrKywadMm9OrVCwBw9+5dVK9eHWfPns3X5KExMTEwNTVFdHQ0TExMiit8IipFZu+/VKj9p3f0LKJICqa0x09EWcvPNQtHbBSTVLUasw9cZlKDyoTEo6eKZQlYOSUmJma6RWX9+vWIiIiAt7d3kfY1YMAAODo6Yvny5QXaf/78+Rg9ejSTGmXcpEmTcPLkSQQHB+PMmTPo0aMHtLW10a9fP5iammL48OGYMGECjh8/jkuXLmHo0KHw8vLiiihERERU7nGOjWLy14VA3H0eJXcYREUmZtkqWP6+CFoGBb/XvyQ5d+4cxo8fj969e8PS0hKXL1/G6tWrUatWLfTu3btI+9LS0sKNGzcKvD9XvSgfnjx5gn79+iE8PBxWVlZo1qwZzp07BysrKwDA4sWLoaWlhV69eiEpKQnt27fPNDEuERERUXnExEYxCI2Oxwr/W3KHQVSk1K/CEbf2L5iM+ljuUIqEs7MzHBwcsHTpUkRERMDCwgKDBg3CDz/8AJVKJXd4VA5t3rw5x+16enpYvnx5gUf+EBEREZVVTGwUg7kHr+B1SprcYRAVudf7jkC/U1soKzu9036HDBmCIUOGFGmbzs7O8PPzK9I2iYiIiIjo3eMcG0Xs8J0n8H8QJncYRMVDrUbsb75yR0FERERERCThiI0i9DolFT8euSp3GETFKuXaTST6n4Nec05YSERERKUfV1UhKv04YqMIrT9/Dy9iX8sdBlGxi/1jA0RystxhEBERERERMbFRVMLjE+F7/p7cYRC9E+oXLxG/lfNTEBERERGR/JjYKCK/+d9CQnKq3GEQvTPxW3ZBHRUtdxhERERERFTOcY6NIvAkMg7brwbJHQbRu5WUhPh/dsF4xCC5IyEiIiIZFXaOCiKiwuKIjSKw4t9bSFULucMgeucS9h7mqA0iIiIiIpIVExuF9PBVDPbdDJE7DCJ5/P+oDSIiIiIiIrnwVpRCWnvuLjhYg8qzhL2HYdi7O7TMTOUOhYhKkJilKwu0n8nYT4o4EiIiIirrOGKjEF7Gvcb+W4/lDoNIXhy1QUREREREMmJioxA2XryPlDS13GEQye71vsNQx8XLHQYREREREZVDTGwUUEJyKrZeeSh3GEQlgkhMwuvDx+UOg4iIiIiIyiHOsVFAO64GISYxRe4wSCZ3YqOx4P4tXIuJxMukROhra6OqkQk+r1wN7aztpXqXoyLw99NgXIkOx63YaKQKgbAOvfPcz5IHt3HwRSiCE+IQn5YKez0D+FjZ4Ysq7qig0pXqRackY8qtKzj68hlMlSqMq+KO/pUqa7QVEB2BHudP4ESz9nAyMCz8SXjL690HYdC9ExRazJcSEREREdG7w08gBZCmFvjr4n25wyAZPUlMQHxaCvpUdMJ37u9hfJUaAIBBl09jw+P/jeQ5+vIZNj55CEABJ/38JxOuxUSilokZxlVxx7wa9dDe2h6bnwah67ljiE9NlerNunsNZyJe4EvXmmhrZYeJNy7iQuQrabsQAtNuB2CEs1uxJDUAIO3ZcyRfuFIsbRMREREREWWHIzYK4OT9UDyN4nwC5ZmPlR18rOw0yoY5uaLdmcP4PfgeBjq4AACGOFbBaJfq0NfWxpRbl/EgJC5f/ayu2yRTWX0zS3wccBaHX4bifTtHAMDhF6H4tpoH+lR0BgDcio3GoRfP0MC8AgBg27MQPHkdjy9cWuT3UPMlwe8AdBt5FmsfREREREREGXHERgHsuhYsdwhUAmkrFLDXM0B0SrJUZqWrB31t7SLtx+H/R35Ep/zvVqhEdRrMlCrpsblKhdfqNyM64lNTMefudUytWhuGOsWby0y+cg2pT0KLtQ8iIiIiIqKMOGIjnyISkvDvgzC5w6ASIj41FYnqNMSmpuDgi1AcexWG7rYORdqHEAIRKclIEwIP42Mx5951aCsUaGJhJdV5z9QCvwXfg6uhMR4lxOP4yzAsqPVm5MTSh7dhq6eP3vZORRpXNsHi9f4jMB4xqPj7IiIiKiFm779UqP2nd+RoRyKiwmBiI5/23wxBqlrIHQaVELPuXsX6/59TQwtAJ5tKmFujbpH28TI5CR7Hd0uP7fX08atHI7gZmUhls6u/hwGX/NHE/wAAoLNNRfSwc8SjhHisDA7E1oYtoVAoijSu7CSePA2j4R9xElEiIiIiInonmNjIJ7/rj+QOgUqQEU5u6GJTCWFJr+EX9gRqCKSo1UXah5lShS31WyBRnYYbMVHY9/wp4tNSNeq4G5viTPOOuBMXDVMdFSobGgF4k3jpbFsRnmaW2Bv2BAsf3EJsago+rFgZE6q4F0uyQx0eieSrN6Fbt3aRt01ERERERPQ2Jjby4d6LKNx5HiV3GFSCuBmZSCMn+lR0Rt8LpzDw8r/Y37hNkSUNVFpaaFHBBgDQztoezS2t0fX8cVRQ6WosLaunrY33TC2kx/+Gv8CJV2E43bwj7sfF4rOr5/BjTU846Bti5NVzsNfTR7+3loQtKonH/ZnYICIiIiKid4JjxfOBozUoN11sKyEgOhIP4vO3+kl+NDCvABtdPWwPDcm2TpoQmHb7Csa4VIednj78wh6jvlkF9KtUGc0srTHQoQq2P8t+/8JKOn0eIjk594pERERERESFxMRGPhy8/UTuEKiES0xLAwDEpKbkUrNwktRpOfbhG/IAcampGOlcDQAQlvQatnp60nZbXT2EJb4utvhEwmsknb9cbO0TERERERGl460oeXQ7LBIvYovvgyCVLi+TEmGlq6dRlqJW45/QYOhraaNahok98yowLgb62jqopG8A4M2KKwoFYKCt+TbdE/YEUSkpqGNqnmU7kcnJWHD/JubXrAe9/19q1kqlhyvREf/rKz42U/xFLfHkaeg1b1ysfRCVJPHx8Th58iRCQkKQ/NaIpbFjx8oUFREREVHZx8RGHp28/0zuEKgE+ermJcSmpqKxRQXY6erjRXIitoeGIDA+FjOr1YGhzpu31uPX8dga+uYWpqvRkQCAxQ9uAQAq6Rmid8X/LcHa/N+D8DK3wo5G3gCAoIQ49LlwEt3sHOBqaAwtKHA1JhLbQh/BQd8AI5zcsoxt/v0bqG5kim4Zlp3tYlsJix7cwlc3L8FB3xAbHj/AzOrvFfVp0ZB85RpEaioUOvwzQ2XflStX0KlTJyQkJCA+Ph4WFhZ49eoVDAwMYG1tzcQGERERUTHiJ448OhXIxAb9T3c7B2x8EoR1IQ8QmZIMIx0deJiYY1o1D7TPMKFnSEI85gfe1Ng3/bGXuZVGYuNtdnr66GxTCafDX2DL02CkqgUq6RtgmKMrvqjiDguVbqZ9bsdGY+PjIOzzaqNR7m5sip9rN8CC+zcRn5qKwY6uGOjgUphTkCuR8Bopt+5C5VGzWPshKgnGjx+Prl274rfffoOpqSnOnTsHpVKJjz76CF988YXc4RFRGTd7/yW5QyAikhUTG3nwKi4Rt8Ii5Q6DSpD37Rzxvp1jrvWaWlojrEPvPLX5dj1LlS5+quWZr7jcjU0R0r5Xltv6VnRG34rO+WqvsJIuXGFig8qFgIAA/P7779DS0oK2tjaSkpLg4uKCH3/8EYMHD0bPnj3lDpGIiIiozOLkoXlw6sEzCLmDICqFki9ekTsEondCqVRCS+vNf6nW1tYICXmz6pCpqSkeP34sZ2hEREREZR5HbOTBKc6vQVQgqcGPkfbyFbStKsgdClGxqlu3Li5cuAA3Nze0bNkS06dPx6tXr7BhwwbUqlVL7vCIiIiIyjSO2MiFWghcePRC7jCISq2kCwFyh0BU7ObOnQs7OzsAwJw5c2Bubo6RI0fi5cuXWLlyZYHa/OGHH6BQKDBu3DipLDExEaNGjYKlpSWMjIzQq1cvPH/+vCgOgYiIiKjU4oiNXNx/GY24pFS5wyAqtVJu3gY6+cgdBlGxql+/vvS7tbU1Dhw4UKj2Lly4gN9//x0eHh4a5ePHj8fevXvxzz//wNTUFKNHj0bPnj1x+vTpQvVHREREVJpxxEYurj6NkDsEolIt5fY9uUMgKlXi4uIwYMAArFq1Cubm5lJ5dHQ0Vq9ejUWLFqF169bw9PTE2rVrcebMGZw7d07GiImIiIjkxREbubj6JFzuEIhKtbRnz6GOioGWmYncoRAVqXr16uHo0aMwNzdH3bp1oVAosq17+fLlPLc7atQodO7cGT4+Pvj++++l8kuXLiElJQU+Pv8bAVW9enU4Ojri7NmzaNy4cZbtJSUlISkpSXocExOT51iIiIiISgMmNnJx9SkTG0SFlXznHvQa18+9IlEp0r17d+jq6gIA3n///SJpc/Pmzbh8+TIuXLiQaVtYWBhUKhXMzMw0ym1sbBAWFpZtm/PmzcOsWbOKJD4iIiKikoiJjRxEJCQhJDJO7jCISr2U20xsUNkzY8aMLH8vqMePH+OLL77A4cOHoaenV+j20k2ZMgUTJkyQHsfExMDBwaHI2iciIiKSG+fYyME1jtYgKhKcZ4PKugsXLuD8+fOZys+fP4+LFy/mqY1Lly7hxYsXqFevHnR0dKCjo4OTJ09i6dKl0NHRgY2NDZKTkxEVFaWx3/Pnz2Fra5ttu7q6ujAxMdH4ISIiIipLmNjIwa1nkXKHQFQmpN5/KHcIRMVq1KhRePz4cabyp0+fYtSoUXlqo02bNrh+/ToCAgKkn/r162PAgAHS70qlEkePHpX2uXv3LkJCQuDl5VVkx0JERERU2vBWlBwERcTKHQJRmSBeJyLtZTi0rSzlDoWoWNy6dQv16tXLVF63bl3cunUrT20YGxujVq1aGmWGhoawtLSUyocPH44JEybAwsICJiYmGDNmDLy8vLKdOJSIiIioPGBiIwfB4UxsEBWV1MdPmdigMktXVxfPnz+Hi4uLRvmzZ8+go1N0/9UuXrwYWlpa6NWrF5KSktC+fXv8+uuvRdY+ERERUWnExEY21EIgJIIThxIVlbTHT4F6HnKHQVQs2rVrhylTpmDXrl0wNTUFAERFRWHq1Klo27Ztgds9ceKExmM9PT0sX74cy5cvL0y4RERERGUKExvZCI1OQGJqmtxhEJUZqY+fyh0CUbFZsGABWrRoAScnJ9StWxcAEBAQABsbG2zYsEHm6IiIiIjKNiY2shEcHiN3CERlChMbVJZVrFgR165dw19//YWrV69CX18fQ4cORb9+/aBUKuUOj4iIitHs/ZcKvO/0jp5FGAlR+cXERjaCOL8GUZFKY2KDyjhDQ0N88skncodBREREVO4wsZGNJ1HxcodAVKaoo6IhUlKhUPLPDpVNgYGBOH78OF68eAG1Wq2xbfr06TJFRURERFT28RNGNsLjE+UOgahsEQLq8Aho21rLHQlRkVu1ahVGjhyJChUqwNbWFgqFQtqmUCiY2CAiIiIqRkxsZCM8PknuEIjKnLRXTGxQ2fT9999jzpw5mDx5styhEBEREZU7WnIHUFJFMLFBVOTUkZFyh0BULCIjI9G7d2+5wyAiIiIql5jYyEZ4Am9FISpq6miuNkRlU+/evXHo0CG5wyAiIiIql3grShZS0tSITUyROwyiMkcdFS13CETFwtXVFd9++y3OnTuH2rVrZ1ridezYsTJFRkRERFT2MbGRhYgE3oZCVBzUMXFyh0BULFauXAkjIyOcPHkSJ0+e1NimUCiY2CCiHM3ef0nuEIiISjUmNrIQxcQGUbEQyclyh0BULIKCguQOgYiIiKjc4hwbWUhKTZM7BKKyiYkNKuOSk5Nx9+5dpKamyh0KERERUbnBxEYWUtVC7hCIyiSO2KCyKiEhAcOHD4eBgQFq1qyJkJAQAMCYMWPwww8/yBwdERERUdnGW1GykKpWyx0CUZkkkjgpL5VNU6ZMwdWrV3HixAl06NBBKvfx8cHMmTPx9ddfyxgdERGVVIWdX2V6R88iioSodGNiIwupaUxsEBUHjtigsmrnzp34+++/0bhxYygUCqm8Zs2aePDggYyREREREZV9vBUlC7wVhaiYMLFBZdTLly9hbW2dqTw+Pl4j0UFERERERY+JjSykcMRGqXKwcg2kOjrIHQYRlWP169fH3r17pcfpyYw//vgDXl5ecoVFREREVC7wVpQspHHERqmyLNkQy5y90K2qJ/qFh8Dy+jWAKxKUTEql3BEQFYu5c+eiY8eOuHXrFlJTU7FkyRLcunULZ86cwcmTJ+UOj4iIiKhM44iNLOhoc9hwaeSXrEI/Y1d80vx9XG/RCsLSQu6Q6C0KJjaojGrWrBkCAgKQmpqK2rVr49ChQ7C2tsbZs2fh6cmJ3YiIiIiKE0dsZEFXR1vuEKgQgtO0MBFWUNZsg6Gq1+j8JBD6d+7KHRYBgJJ/cqjsqlKlClatWiV3GERERETlDj9lZEGPiY0yIQUKrEw2wErrOmhUsRZGxIfC8VoAkPBa7tDKLY7YoLIqJCQkx+2Ojo7vKBIiIiKi8oeJjSzo81vlMud8ijbOqxxg1rASxmjFoMm9G9B+8lTusModBd9bVEY5OzvnuPpJWlraO4yGiIiIqHzhp4wsGOrytJRVUWoFvlObAi5N0dM9GX1fBsP8xnUglR863gWFrkruEIiKxZUrVzQep6Sk4MqVK1i0aBHmzJkjU1RERERE5QM/wWfBSJfD5cuD7UkqbDepCrcWrvg85RVq3ggAIqPkDqtMU5iYyB0CUbGoU6dOprL69evD3t4eP/30E3r27ClDVERERETlAxMbWTBmYqNcCUzVwniFNVS122KEKgHtQ+5B716g3GGVSVpmpnKHQPROVatWDRcuXJA7DCIiIqIyjYmNLKh0tGGkq4O4pFS5Q6F3KBkKLE82xHLbumjm6IHhsU9R8WoAkJgod2hlhpY5ExtUNsXExGg8FkLg2bNnmDlzJtzc3GSKioiIiKh8YGIjGzbGBohLism9IpVJ/yZr419dR1g2dsAYRTQa370OrdBncodV6mmbmckdAlGxMDMzyzR5qBACDg4O2Lx5s0xREREREZUPTGxkw87EAA9eMbFR3oWrFZgJM8C1OfrUTELv50EwvXkT4AoHBcIRG1RWHTt2TCOxoaWlBSsrK7i6ukJHh//VEhERERUnLbkDKKlsTQ3kDoFKmC1JuuhtVh1jvHvgTrMWAOeLyDfOsUFllbe3N1q2bCn9NG/eHNWrV89XUmPFihXw8PCAiYkJTExM4OXlhf3790vbExMTMWrUKFhaWsLIyAi9evXC8+fPi+NwiIiIiEoVJjayYWeiL3cIVELdTdHCWC1bdPVoh71tOiPJtYrcIZUOuiooTIzljoKoWMybNw9r1qzJVL5mzRrMnz8/T21UqlQJP/zwAy5duoSLFy+idevW6N69O27evAkAGD9+PHbv3o1//vkHJ0+eRGhoKFdbISIiIgITG9myMeaIDcpZEhRYkmKIrvaemOvTE88aNAT0dOUOq8TSsbfLNAcBUVnx+++/o3r16pnKa9asid9++y1PbXTt2hWdOnWCm5sbqlatijlz5sDIyAjnzp1DdHQ0Vq9ejUWLFqF169bw9PTE2rVrcebMGZw7d66oD4eIiIioVGFiIxt2vBWF8uFEsg4G6ztjYONu+M/bB2pbG7lDKnG0K9nJHQJRsQkLC4OdXebXuJWVFZ49y//Ew2lpadi8eTPi4+Ph5eWFS5cuISUlBT4+PlKd6tWrw9HREWfPni1U7ERERESlHWc0y4aDmaHcIVAp9FytwDRYQOHWAv1rJaHH84cwuXkLUKvlDk12OpXs5Q6BqNg4ODjg9OnTqFy5skb56dOnYW+f99f+9evX4eXlhcTERBgZGWHHjh2oUaMGAgICoFKpYPbWykI2NjYICwvLsc2kpCQkJSVJj99empaIiIiotGNiIxs2JgYw1VMhOjFZ7lCoFBIKBf5K1sNf5jVQs1V1fJ74HG7XA4CYWLlDk412RSY2qOwaMWIExo0bh5SUFLRu3RoAcPToUXz11VeYOHFintupVq0aAgICEB0dja1bt2Lw4ME4efJkoWKbN28eZs2aVag2iIiIiEoyJjZyUNXGFBcevZQ7DCrlbqZoYZS2HQzq2uIznXi0DroN1cMgucN65zhig8qyL7/8EuHh4fj888+RnPwmIa6np4fJkydjypQpeW5HpVLB1dUVAODp6YkLFy5gyZIl6Nu3L5KTkxEVFaUxauP58+ewtbXNsc0pU6ZgwoQJ0uOYmBg4ODjk4+iIiIiISjbOsZGDatZcmpKKToJQYFGKEbpUaoCf2vbE8/r1AZVK7rDeGW0mNqgMUygUmD9/Pl6+fIlz587h6tWriIiIwPTp0wvVrlqtRlJSEjw9PaFUKnH06FFp2927dxESEgIvL68c29DV1ZWWkE3/ISIiIipLOGIjB1WtzeQOgcqow0k6OGzggopNK2O0OgL17lyH4vkLucMqNtoV7aBlyAl5qewLCwtDREQEWrRoAV1dXQgh8rwa0JQpU9CxY0c4OjoiNjYWGzduxIkTJ3Dw4EGYmppi+PDhmDBhAiwsLGBiYoIxY8bAy8sLjRs3LuajIjnFLF2Z731Mxn5SDJGUbbP3X5I7BCIiKgQmNnJQzcZM7hCojHuapsAUWEJRtSUG105Ct2f3YXT7NqAWcodWpJTVXOUOgahYhYeHo0+fPjh+/DgUCgUCAwPh4uKC4cOHw9zcHAsXLsy1jRcvXmDQoEF49uwZTE1N4eHhgYMHD6Jt27YAgMWLF0NLSwu9evVCUlIS2rdvj19//bW4D42IiIioxGNiIwdVKphAR0uB1DL2IZNKHqFQwDdZD76WtVCnVQ2MfB0Gl+sBQGyc3KEVCWU1N7lDICpW48ePh1KpREhICNzd3aXyvn37YsKECXlKbKxevTrH7Xp6eli+fDmWL19e6HiJiIiIyhImNnKg1NaCq5Up7jyPkjsUKkeupmjhMx17GNWzw+facWj58BaUwY/kDqtQOGKDyrpDhw7h4MGDqFSpkka5m5sbHj0q3e9fIiIiopKOiY1c1He0YmKDZBEnFPgx1Rg/OjZCJ9d6+CgyBBWuXQNSUuQOLX+USui4OMsdBVGxio+Ph4FB5nlkIiIioKurK0NEVNIUZK4MIiIiyhuuipKLBk5WcodAhH3JSvQ3rIKPm3VHQMvWEFYV5A4pz5QuzlAomUOlsq158+ZYv3699FihUECtVuPHH39Eq1atZIyMiIiIqOzjp41c1He0grZCgTTBeTZIfiFpWvgKFaDt3gpD6ySi89NAGN65C5Tg16eytnvulYhKuR9//BFt2rTBxYsXkZycjK+++go3b95EREQETp8+LXd4RERERGUaR2zkwkhXieq2ZnKHQaQhDQr8kayPHlYemNq6F4K9mgJGhnKHlSVdz/fkDoGo2NWqVQv37t1Ds2bN0L17d8THx6Nnz564cuUKqlSpInd4RERERGUaR2zkQUMna9x8FvlO+kp4+hDPj25BwqM7UKemQGVuA8uGPqjQpBMAQKjViLhwBOH/HUJyeBi0lHrQr1gZ1q0+gKFTtRzbjrh0HE+2Zb80oEOfsTB/rzkAIP7RHTzdvQbJr55Bv6ILKnb/BHrWFTXqP929BkmvQuEydFohj5oK42KKFi4qK8Kkvj1GacWi+f2b0Al5LHdYAACFni6UNavLHQZRsUpJSUGHDh3w22+/4ZtvvpE7HCIiIqJyh4mNPGjgZIW15+4Wez+xgVcRvP4H6Nu/SVRo6eohOTwMKdHhUp1nBzbg1b97YPZec1g2ao+0xARE/HcYD1bNgOun38HAIftlNQ0r14BD7zGZyl+d3oPXYY9gVKUWACAtMR7BG36EgYMbLBv4IPLyCTzauABVxy6AQksbAJD4/DEiLhyB26j5RXwWqKBi1ArMU5tgnrMXulX1RL/wEFhevwakpsoWk9KjJufXoDJPqVTi2rVrcodBREREVG7xE0ce1HOoAF0dLSSlqoutj7TEBDz+ZxmMq9WDU/+JUGhlvktIpKUh/PwhmNZqDMc+Y6Vys1qNcWfBaEQG+OeY2NC1sIGuhY1GmTolCU/9/oCRSy0ojc0BAAkh96BOSYZT/4nQUqpgXPU93PlpFJLCw6Bn9WbURujetbBo4AM9G4eiOHwqYn7JKvgZu8K5uQvGpIWj1s2rUIRHvPM4dOu/9877JJLDRx99hNWrV+OHH36QOxQiIiKicoeJjTzQV+qgSWVbHA8MLbY+oq7+i9S4aNi26weFlhbUyYlQ6Kg0EhxCnQqRkgwdI1ONfXWMTAGFAlpKVb77jbl9Ceqk1zD7/1tQAECdkgwtHaXUnra+0Zv+U5IBANG3/sPr0GA49puQ7/7o3QpO08JEWEFZsw2Gql6j85NA6N8p/tFH6VT1676zvojklJqaijVr1uDIkSPw9PSEoaHmnDeLFi2SKTIiIiKiso+JjTzyqV6xWBMbcfevQUtXHykxEQj+80ckv3oGLZUuzN5rAfvOQ6ClVEFLqQsDBzdEXj4BA8dqMHSujrTXCXhxfCu09Y1g0bBtvvuNuuoPhVIF05qNpDJ9+8pIS0rAS//dMK3VGK/O7IWWngF0K9hDnZqCZ/vWw6ZNH+j8f8KDSr4UKLAy2QArreugUcVaGBEfCsdrAUDC62LrU7uSPXTsbHKvSFSKPXz4EM7Ozrhx4wbq1asHALh3755GHYVCIUdoREREROUGExt55O1mD6W2FlLSiud2lKTwMAi1GsEbfoRF/dYwajcAcUE3EX52P9ISE+D04TgAgEOfMQjZtBiPtyyV9lVZ2KDKp99lus0kN6kJsYi9FwCTGg2grav/v/bMrWHXfgCeHfwTz/avh0KpQqWeI6Gl0sWLE9uhpVTBslH+kyhUMpxP0cZ5lQPMGlbCGK0YNLl3A9pPnhZ5P3otmhR5m0QljZubG549e4bjx48DAPr27YulS5fCxoZJPSIiIqJ3hYmNPDLSVaKxszX8H4QVS/vq5ESIlCRYNGyLil2HAQBMazWCSEtFxH+HkeTTF7oV7KCl0oeujQMMHKvCqEptpMZG4cWpnXj050+o8sls6Bia5LnP6BvnINJSNW5DSWfVvBvM3muB5MgX0LWyh46+EVJiIvDixA44ffQVhFqN0D2+iLl9ATrGZrDvPASGTlz9ojSJUivwndoUcGmKnu7J6PsyGOY3rgOpaUXSvp530yJph6gkE0JoPN6/fz/i4+NlioaIiIiofMo8QyVlq231SsXWtkLnzXwWZnWaaZSnP04IuQeRloagNbOhrWuAit0+hmnNRrBs3B4uw6YjOSIML/398tVnVIA/tPWNYFI163kQlMZmMHSsKt1y8uzAnzCqUhvGrrXx4thWxD24Dsd+42FaoyGC1s1D2mtezJdW25NU6GtSFaNa9MDN5t6AuVmh2tNxcYaOQ8XcKxKVMW8nOoiIiIio+HHERj60crOHjpYCqeqiv3BVmpgj6cXjzBODGr55nPY6DvHBt5D4/DHsOg3WqKNbwQ66VpUQ/+hOnvtLjnqJ+Ed3YNHABwrt3F8G8SH3EH3jHKp+8WYCvKhrp2HdqhcMHavB0LEawv87jJg7l2Bet0WeY6CSJzBVC+MV1lDVbosRqgS0D7kHvXuB+W6HozWovFAoFJnm0OCcGmVbzNKVcodAREREb2FiIx9M9FVo6mKLk/efFXnb+hVdEHf/GlJjIgCr/33TnRr7ZolObUMTpMRFvykUmef5EGmpgDrv839EXT0NCAHzOplvQ8nUthAI3bMGFZp0gq6lLQAgJSYCShMLqY7S2BwpMe9+OVEqHslQYHmyIZbb1kUzRw8Mj32KilcDgMTE3HdWKKDXkokNKh+EEBgyZAh0dXUBAImJifjss88yrYqyfft2OcIjIiIiKhd4K0o+9XqvcrG0a1b7zUSLERePaZRHXDgKaGnDyKUmdCvYAXgzWiKjhKcPkfQqFHp2zlKZOjkJiS+eIjU+Jsv+oq7+C6VZBRg45z4vRuTlE0iJDod1q15SmY6RGZJevplwUqSlIik8DDrGZrm2RaXPv8naGKrriH6Nu+JMq3ZQ29vlWF/pXhXa1hXeUXRE8ho8eDCsra1hamoKU1NTfPTRR7C3t5cep/8QERERUfHhiI18au5qBzsTAzyLSSjSdvXtK8PcsxUiLx2HUKfBsHJNxD+8iegbZ2HVsgeUJhZQmljAyNUDkZdPIi3xNYzd6iAlNhLhZ/dDS6lChaadpfYSntzHwz9mwrp1b9j69NHoKzEsBIlhj2DV8v1ch0ynJb1G2KGNsG3XX2PlFNNajfH82FYIIZDw6A5Eakq2c3VQ2RCuVmAmzADX5uhTMwm9nwfB9OZNIE1zslH9Dm3kCZBIBmvXrpU7BCIiIqJyj4mNfNJSKNDzvcpYfupmkbdd6f1PoDKzQsSl44i59R+UZlaw6zwEVhkSFs4Dv8JL/92IunYasYEBUGjrwNDZHbY+faFnlbfJGiOv+gPIPFFpVl4c2wqliSXM63lrlNv69EFafAxeHP0HOsZmcOo/MdP8IFR2bUnSxRaz6qjmXRWjkl6g+o2rQFQ0FEaGXOaViEoNzpdRdszef0nuEIiISEYKwSnc8+1VXCLaL99bLJOIEpVGuhD4TJmAdg5mqNS3u9zhEFEOYmJiYGpqiujoaJiY5H2J8Hz3U8CkgcnYT4o4kuyV1cTGuzyHJQUTG1ReTe/oKXcIRMUmP9csnGOjACoY6cHbzV7uMIhKjCQosDTFEGltW8sdChERERERlTNMbBTQB3Vd5A6BqERpVsUWThbGcodBRERERETlDBMbBdTY2RpVrTmnBFG6/vVd5Q6BiIiIiIjKISY2CkihUODjJrkvlUpUHlSzMYNXZRu5wyAiIiIionKIiY1CaFu9Eipbcug90adN3XNdOpiIiIiIiKg4MLFRCFoKBYZ7cdQGlW/VbMzQuion0yUiIiIiInkwsVFInWo6opKZodxhEMmGozWIiIiIiEhOTGwUkraWAsO8qskdBpEsOFqDiIiIiIjkxsRGEehW2xn2pgZyh0H0znG0BhERERERyY2JjSKg1NbCuFa15Q6D6J2qztEaREVq3rx5aNCgAYyNjWFtbY33338fd+/e1aiTmJiIUaNGwdLSEkZGRujVqxeeP38uU8REREREJYOO3AGUFe3dHbDp4n1ceRIudyhE78TktnU4WoOoCJ08eRKjRo1CgwYNkJqaiqlTp6Jdu3a4desWDA3fzOU0fvx47N27F//88w9MTU0xevRo9OzZE6dPn5Y5epLbpccvpd/991/K9/7TO3oWZTj5NrsAMRMREaVjYqMIfeXzHvr7HoWQOxCiYtaxhgPqOVjJHQZRmXLgwAGNx76+vrC2tsalS5fQokULREdHY/Xq1di4cSNat24NAFi7di3c3d1x7tw5NG7cWI6wiYiIiGTHW1GKUA07c3St7SR3GETFSl+pjfGtPeQOg6jMi46OBgBYWFgAAC5duoSUlBT4+PhIdapXrw5HR0ecPXtWlhiJiIiISgImNorYWO9aMFBxIAyVXR83qQ4bY325wyAq09RqNcaNG4emTZuiVq1aAICwsDCoVCqYmZlp1LWxsUFYWFi2bSUlJSEmJkbjh4iIiKgs4SfwImZlpI/hXtWw7ORNuUMhKnIOZoYY1LCq3GEQlXmjRo3CjRs38O+//xa6rXnz5mHWrFlFEFXJFbN0pdwhlCjNjx3IvdJbYgIvwWTsJ8UQDRGVZIWd30bu+XmI0pWbERuPHz+Gnp5egSZYu3XrFnR0dHDjxo081R/cqBqqVDDJdz9EJd0knzpQ6WjLHQZRmTZ69Gjs2bMHx48fR6VKlaRyW1tbJCcnIyoqSqP+8+fPYWtrm217U6ZMQXR0tPTz+PHj4gqdiIiISBb5Smz4+vpCoVBIP3p6eqhatSpGjx5d4pebmz17Nho1aoSmTZtqlD99+hR9+vSBmZkZTExM0L17dzx8+FCjTo0aNdC5c2dMnz49T30ptbUws5MntLhgBJUhHdwd4O3G5V2JiosQAqNHj8aOHTtw7NgxVK5cWWO7p6cnlEoljh49KpXdvXsXISEh8PLyyrZdXV1dmJiYaPwQERERlSUFuhVl9uzZqFy5MhITE/Hvv/9ixYoV2LdvH27cuAEDA4OijrHQXr58iXXr1mHdunUa5XFxcWjVqhWio6MxdepUKJVKLF68GC1btkRAQAAsLS2lup999hk6deqEBw8eoEqVKrn26VHREgMauGHDf4FFfjxE75qVkR6mtq8rdxhEZdqoUaOwceNG7Nq1C8bGxtK8GaamptDX14epqSmGDx+OCRMmwMLCAiYmJhgzZgy8vLy4IgoRERGVawVKbHTs2BH169cHAHz88cewtLTEokWLsGvXLvTr169IAywKf/75J3R0dNC1a1eN8l9//RWBgYH477//0KBBAwBvjq1WrVpYuHAh5s6dK9X18fGBubk51q1bh9mzZ+ep39EtasH/fhiCI2KL7mCIZDCjoydM9VVyh0FUpq1YsQIA4O3trVG+du1aDBkyBACwePFiaGlpoVevXkhKSkL79u3x66+/vuNIiYiopCjsHBlEZUWRzLHRunVrAEBQUBAAYMGCBWjSpAksLS2hr68PT09PbN26NdN+r1+/xtixY1GhQgUYGxujW7duePr0KRQKBWbOnCnVe/ToET7//HNUq1YN+vr6sLS0RO/evREcHJyn+Hbu3IlGjRrByMhIo3zr1q1o0KCBlNQA3iyd16ZNG2zZskWjrlKphLe3N3bt2pWnPgFAT6mN77rUh7aC96RQ6dWjjjOau9rJHQZRmSeEyPInPakBAHp6eli+fDkiIiIQHx+P7du35zi/BhEREVF5UCSJjQcPHgCAdOvGkiVLULduXcyePRtz586Fjo4Oevfujb1792rsN2TIECxbtgydOnXC/Pnzoa+vj86dO2dq/8KFCzhz5gw+/PBDLF26FJ999hmOHj0Kb29vJCQk5BhbSkoKLly4gHr16mmUq9VqXLt2TRp5klHDhg3x4MEDxMZqjrTw9PTEjRs38rVUnkdFSwzzqpbn+kQlib2pAb5sU0fuMIiIiIiIiLJVoFtRoqOj8erVKyQmJuL06dOYPXs29PX10aVLFwDAvXv3oK+vL9UfPXo06tWrh0WLFkmJi8uXL2PLli0YN24cFi9eDAD4/PPPMXToUFy9elWjv86dO+ODDz7QKOvatSu8vLywbds2DBw4MNtYQ0JC8Pr160yTsEVERCApKQl2dpm/iU4vCw0NRbVq/0tKuLi4QK1W486dO2jYsGGu5yndZ81r4ELISwQ8Cc/zPkRy01IAszvXh6GuUu5QiIiIiIiIslWgxIaPj4/GYycnJ/z111+o+H/t3XtYVWXe//HP3pyRs8gpAUEQ0RBPZHg+gEDGaDlPZk5ho/artHLUTmOplf1sbPRnB8uyZzSnxuyk9VhaionWEE8wkpVGSShOecoTB0EE1u+PYk9bURGBzcb367r2dbHudR++a9+tLvi67ntddZUkWSU1jh8/rpqaGg0aNEirV6+2lG/c+Ms71u+++26rvu655x6tXLnSquy3/Z05c0YlJSWKioqSj4+P/vWvf10wsXH06C/JBF9fX6vyiooKSb/sFn82V1dXqzp16vr4+eefzztefRzNZj095lqN+9tmHTt1+pLaArbyfwZ2U0J4gK3DAAC0kJJnX25UO69772CdPwDAphqV2Fi6dKm6dOkiR0dHBQYGKiYmRmbzf1a1rF+/XvPnz1d+fr5On/7PH/Km3+w1sW/fPpnN5nOepIiKijpnvIqKCi1YsEArVqzQjz/+KMMwLOdOnjzZoJh/20b6T7Lkt/HVqaystKpzdh+mRuyZEeDppqdG99Ndb2xXzVmxAK3N4Khg/Z8BsbYOAwAAAAAuqlGJjWuuuabevSkkafv27frd736nwYMH64UXXlBwcLCcnJy0YsUK/eMf/2hUkPfcc49WrFih6dOnKzExUd7e3jKZTLr55ptVW1t7wbZ1+34cP37cqtzPz08uLi46cODAOW3qykJCQqzK6/rw9/dv1HX06xSguwZ10/PbvmlUe6AlhPq005PpCY1K4AEAAABAS2tUYuNC3nnnHbm6uuqjjz6yWuaxYsUKq3rh4eGqra1VUVGRoqOjLeV79uw5p8+3335bGRkZWrRokaWssrJSJ06cuGg8YWFhcnNzs7yxpY7ZbFZcXJxyc3PPaZOTk6PIyEh5enpalRcVFclsNqtLly4XHfd8Jvfvqi9/PKrthQcb3QfQXFydHPT/xvaXlyuvdgUAAABgH5o8seHg4CCTyaSamhpL2d69e7Vu3TqreikpKZo9e7ZeeOEFy+ahkvTcc8/V2+fZS0mee+45qzHOx8nJSX379q03gfH73/9eDz30kHJzcy1PoBQUFGjLli2aNWvWOfXz8vLUvXt3eXt7X3Tc8zGZTPq/6ddo/MpM/ftEeaP7AZrD3LQ+ig5o/H/fAHC5GrvPAwAAuHI1eWJj1KhRWrx4sVJTU3XLLbfo8OHDWrp0qaKiorRz505LvT59+mjs2LFasmSJjh49qmuvvVZZWVn67rvvJFnvY3H99dfr73//u7y9vdWtWzdlZ2dr8+bNlmUmFzN69GjNnj1bJSUl8vLyspTffffdWr58uUaNGqVZs2bJyclJixcvVmBgoGbOnGnVx5kzZ5SVlXXOZqeN4eXmrKU3DdRtqz7Rycqqy+4PaAp/SIjWdd3DbB0GAAAAAFwS88WrXJrhw4frv//7v3Xw4EFNnz5dq1ev1l/+8hfdcMMN59RdtWqVpk6dqg8++EAPPvigqqqqtGbNGkn/eTOJJD3zzDO67bbb9Prrr2vmzJk6cOCANm/eLA8PjwbFdOutt6qmpkbvv/++Vbmnp6e2bt2qwYMHa/78+Xr00UcVHx+vrKwsdejQwapuZmamjh07poyMjEv9SurVqb2nnvmv/nJxbPIpAC7ZyNiOmjmih63DAAAAAIBLZjLOXuNhY/n5+erVq5dee+01TZgwocn6nTRpkr777jtt3769Ue3HjBkjk8mktWvXNllMkrT523/r/nWfq7ZVzQKuJAnhHfTiuEFyciDJBlwJSkpK5O3trZMnT1o9xdjk47CkpEXl7T9yWe37hHa4eKXz4HWvwJVrTlofW4eANuxSfmdp8qUol6KiouKcV6ouWbJEZrNZgwcPbtKx5s6dqy5duuizzz7TgAEDLqnt7t27tX79euXn5zdpTJKU1LWjZo2I18LNXzZ538DFxAR4a8nY/iQ1AACNVvLsyxp0iYmV7cNTmykaAC3J1klNEiuoY9PExsKFC5WXl6dhw4bJ0dFRGzZs0IYNG3THHXcoNDS0SccKCwtTZWVlo9rGxsaqurq6SeP5rQkJ0frp5Cm99sX3zTYGcLYQb3ctHTdQHi5Otg4FAAAAABrNpomN/v37a9OmTXriiSdUVlamsLAwzZs3T7Nnz7ZlWDYxa0QPlVed0dov99o6FFwBfN2c9eK4Qerg4XbxygAAAADQitk0sZGcnKzk5GRbhtBqmEwmzU3rI7PJpHfyi2wdDtowP3cXvTR+kDq197R1KAAAAABw2Wya2IA1k8mkR1N7SxLJDTSLDh6uenn8YEX6N9+GgQAAAADQkkhstDJ1yQ0Hk0lv7vjB1uGgDQnyctPy8UMU5tew1yQDAAAAgD0gsdEKmUwmzU7tLZPJpDX/KrR1OGgDQrzdtfyWIero087WoQAAAABAk+Idj63Yn1N6aVJijK3DgJ0L8/XQij8MJakBAAAAoE0isdHK3Ts0TnPT+sjRbLJ1KLBDPa9qr1dvG6YgL3dbhwIAAAAAzYLEhh24sWeEnvuvAfJwYeUQGu66bqFafstg+bm72DoUAAAAAGg2JDbsRP/IIK38wzAFebnZOhTYgbsGddOC0f3k7Ohg61AAAAAAoFmR2LAj0QHe+vttwxUb5GPrUNBKuTia9dTofrpzYDdbhwIAAAAALYK1DXYmwNNNK/8wTAs+3qF1O/faOhy0Iv7tXLX4xkTFd2xv61AAALigQVs2Nqrd9uGpTRwJAKAt4IkNO+Tq5KDHRvXV/OsT5OrEUgNI14R30JuTkkhqAAAAALjikNiwY+lx4Xo9Y7gi23vaOhTYiNkk3TEgVi+NH6z27VxtHQ4AAAAAtDiWoti5qA7een3iCD350b+0/utiW4eDFhTo6aYn0xOUEB5g61AAAGgRjVnCwvIVAGj7eGKjDXB3dtST6ddo/vUJ8nJ1snU4aAHDu4TorUnJJDUAAAAAXPF4YqMNSY8LV2JEoOZ/9C998t1Ptg4HzcDX3UUPJsUrrXuYrUMBAAAAgFaBxEYb4+/hqiVj++uj3fu14ON8HT912tYhoYn8Li5cM4f3kI+7i61DAQAAAIBWg6UobVRKbKjWThmp67qF2joUXKaOPu300s2D9MT1CSQ1gDZu27ZtSk9PV0hIiEwmk9atW2d13jAMzZkzR8HBwXJzc1NSUpK+//572wQLAADQSpDYaMN83V20YHQ/Lb1pAG9OsUOOZpMm9uuitycn69qIQFuHA6AFlJeXKz4+XkuXLq33/MKFC/Xss89q2bJlysnJUbt27ZSSkqLKysoWjhQAAKD1YCnKFWBg52BdGxGod3YU6cVPd7E8xQ4M6xKi6UPj1ImEFHBFSUtLU1paWr3nDMPQkiVL9Mgjj2j06NGSpFWrVikwMFDr1q3TzTff3JKhAgAAtBokNq4QjmazxvXprOu6h+mV7N36xxd7VFVTa+uwcJb4q/z0p+E91Kujv61DAdDKFBUV6eDBg0pKSrKUeXt7q1+/fsrOzj5vYuP06dM6ffo/Ce2SkpJmjxUAAKAlsRTlCuPp6qQ/DeuhdXekKDU2VCZbBwRJUpivhxbdcK1W3TacpAaAeh08eFCSFBhovTQtMDDQcq4+CxYskLe3t+UTGsreSwAAoG0hsXGFusqnnf4ypp/empyslNiOMpPhsIkwXw89mtpba+8YqaSuHW0dDoA26OGHH9bJkyctn/3799s6JAAAgCbFUpQrXHQHby0cc62Kjpbob9kF+vCbYlXXGrYOq83rHuSr2xNjNCLmKplNZJUAXFxQUJAk6dChQwoODraUHzp0SD179jxvOxcXF7m48EYlAADQdvHEBiRJEe299MT1CfrwrjTddk202jmT82oOAyIDtfyWwfrH7SOU3LUjSQ0ADRYREaGgoCBlZmZaykpKSpSTk6PExEQbRgYAAGBb/PUKK4Fe7po5Il53Dequjbv26538H/T1geO2Dsuuebg4KiU2VON6d1ZMoI+twwHQipWVlWnPnj2W46KiIuXn58vPz09hYWGaPn265s+fr+joaEVEROjRRx9VSEiIxowZY7ugAQCwU49vyLus9nPS+jRRJLhcJDZQL3dnR93YM0I39ozQd4dP6J38In3wTbFKK8/YOjS7YJLUJ6yDxvTopKSuV8nNiVsNwMXl5uZq2LBhluMZM2ZIkjIyMrRy5Uo98MADKi8v1x133KETJ05o4MCB2rhxo1xdXW0VMgAAgM3x1xYuqkuAjx4e2Ut/GtZDm779t9Z/vU+5xUfYi6MeQV5uSr86XGN6dFJHXw9bhwPAzgwdOlSGcf7/t5pMJj3++ON6/PHHWzAqAACA1o3EBhrM1clB6XHhSo8L18mKKmV9/5Myv/tR2UWHdLq61tbh2Uykv5eGRQdraHSI4kL8ZGLfDAAAAABoMSQ20Cjebs76XY9O+l2PTjpVVa3PfjioLQU/6rMfDulkZZWtw2tWDiaTenZsr6FdQjQ0KkRhfjyZAQCwb3n7j9g6BAAAGo3EBi6bu7Ojkrt2VHLXjjIMQ3uOlCi3+Ijy9h9RXvHPOnbqtK1DvCyOZpO6BfmqV6i/enX0V+9Qf3m7Ods6LAAAAACASGygiZlMJkUHeCs6wFvj+0ZJkoqOlii3+Gd99dMx7Tl8UoU/l6iyusbGkZ5fBw9XdQnwVs9fkxhXB/vJ1cnB1mEBAAAAAOpBYgPNLqK9lyLae+m/ekVKkmoNQ/uPl+n7IyX6/vAJfX+kRHuPlupwWUWLvXXFJCnA002R/l7q7O+pzv7eivT3VKS/l7xceRoDAAAAAOwFiQ20OLPJpHA/T4X7eSop5iqrc6eqqnW4tEKHSit0uLRCh8sqdLSsUqfOVKvyTI0qz9So4ky1Kqt/+fl0dY0czCY5OzjI2dEsZweznB0c5PTrz56uzvJzd/nl085FAZ5uCvBwU4Cnm5wczDb6BgAAQGs3aMvGRrXbPjy1iSMBcD6Pb8izdQhoJUhsoFVxd3ZUp/ae6tTe09ahAAAAAADsAP9kDQAAAAAA7BaJDQAAAAAAYLdIbAAAAAAAALtFYgMAAAAAANgtEhsAAAAAAMBukdgAAAAAAAB2i9e9AgAAAFeQQVs2XnKb7cNTmyESAGgaPLEBAAAAAADsFokNAAAAAABgt0hsAAAAAAAAu8UeGwAAAGizGrOfREuzhxjbMvYcAewfT2wAAAAAAAC7RWIDAAAAAADYLRIbAAAAAADAbrHHBgAAsHt5+49cVvs+oR1sOj7aDvbLsNbY76Mxe1jw3cPePL4h77Laz0nr00SR2D+e2AAAAAAAAHaLxAYAAAAAALBbJDYAAAAAAIDdYo8NAAAAABfU0vtXtNX9MlpyzxE0v8vdI8Pex29Ne3zwxAYAAAAAALBbJDYAAAAAAIDdIrEBAICdWbp0qTp16iRXV1f169dP//u//2vrkAAAAGyGPTYAALAja9as0YwZM7Rs2TL169dPS5YsUUpKigoKChQQEGDr8OxW3v4jtg4BAFqFltwHhD1H0FR4YgMAADuyePFiTZkyRbfffru6deumZcuWyd3dXX/7299sHRoAAIBNkNgAAMBOVFVVKS8vT0lJSZYys9mspKQkZWdn2zAyAAAA22EpCgAAduLnn39WTU2NAgMDrcoDAwP17bff1tvm9OnTOn36tOX45MmTkqSSkpLmC1RSSWVFs/Z/tvLTlS06HoArW+Wpska1a+z/qxo7XmO0ZIz28H3g/Jr9d4lf+zcM46J1SWwAANCGLViwQI899tg55aGhoTaIBgDaiOfmt+3xGqMlY7SH7+MKsKCFxiktLZW3t/cF65DYAADATvj7+8vBwUGHDh2yKj906JCCgoLqbfPwww9rxowZluPa2lodO3ZM7du3l8lkatZ4rzQlJSUKDQ3V/v375eXlZetw8CvmpfViblov5qb1upLmxjAMlZaWKiQk5KJ1SWwAAGAnnJ2d1adPH2VmZmrMmDGSfklUZGZmatq0afW2cXFxkYuLi1WZj49PM0d6ZfPy8mrzv2zaI+al9WJuWi/mpvW6UubmYk9q1CGxAQCAHZkxY4YyMjLUt29fXXPNNVqyZInKy8t1++232zo0AAAAmyCxAQCAHRk3bpyOHDmiOXPm6ODBg+rZs6c2btx4zoaiAAAAVwoSGwAA2Jlp06add+kJbMfFxUVz5849Z+kPbIt5ab2Ym9aLuWm9mJv6mYyGvDsFAAAAAACgFTLbOgAAAAAAAIDGIrEBAAAAAADsFokNAAAAAABgt0hsAAAA/Grbtm1KT09XSEiITCaT1q1bZ3XeMAzNmTNHwcHBcnNzU1JSkr7//nurOseOHdOECRPk5eUlHx8fTZo0SWVlZVZ1du7cqUGDBsnV1VWhoaFauHBhc1+aXbvYvEycOFEmk8nqk5qaalWHeWkeCxYsUEJCgjw9PRUQEKAxY8aooKDAqk5lZaWmTp2q9u3by8PDQ2PHjtWhQ4es6hQXF2vUqFFyd3dXQECA7r//flVXV1vV2bp1q3r37i0XFxdFRUVp5cqVzX15dq0hczN06NBz7p0777zTqg5z0/RefPFF9ejRQ15eXvLy8lJiYqI2bNhgOc89c+lIbAAAAPyqvLxc8fHxWrp0ab3nFy5cqGeffVbLli1TTk6O2rVrp5SUFFVWVlrqTJgwQd988402bdqk9evXa9u2bbrjjjss50tKSjRy5EiFh4crLy9PTz/9tObNm6eXX3652a/PXl1sXiQpNTVVBw4csHxWr15tdZ55aR5ZWVmaOnWqPv/8c23atElnzpzRyJEjVV5ebqnzpz/9Sf/zP/+jt956S1lZWfrpp5904403Ws7X1NRo1KhRqqqq0j//+U+9+uqrWrlypebMmWOpU1RUpFGjRmnYsGHKz8/X9OnTNXnyZH300Ucter32pCFzI0lTpkyxund+m9BjbppHx44d9dRTTykvL0+5ubkaPny4Ro8erW+++UYS90yjGAAAADiHJGPt2rWW49raWiMoKMh4+umnLWUnTpwwXFxcjNWrVxuGYRi7du0yJBlffPGFpc6GDRsMk8lk/Pjjj4ZhGMYLL7xg+Pr6GqdPn7bUefDBB42YmJhmvqK24ex5MQzDyMjIMEaPHn3eNsxLyzl8+LAhycjKyjIM45d7xMnJyXjrrbcsdXbv3m1IMrKzsw3DMIwPP/zQMJvNxsGDBy11XnzxRcPLy8syHw888IDRvXt3q7HGjRtnpKSkNPcltRlnz41hGMaQIUOM++6777xtmJuW4+vra7zyyivcM43EExsAAAANUFRUpIMHDyopKclS5u3trX79+ik7O1uSlJ2dLR8fH/Xt29dSJykpSWazWTk5OZY6gwcPlrOzs6VOSkqKCgoKdPz48Ra6mrZn69atCggIUExMjO666y4dPXrUco55aTknT56UJPn5+UmS8vLydObMGav7pmvXrgoLC7O6b+Li4hQYGGipk5KSopKSEsu/YGdnZ1v1UVenrg9c3NlzU+f111+Xv7+/rr76aj388MM6deqU5Rxz0/xqamr0xhtvqLy8XImJidwzjeRo6wAAAADswcGDByXJ6hfJuuO6cwcPHlRAQIDVeUdHR/n5+VnViYiIOKePunO+vr7NEn9blpqaqhtvvFEREREqLCzUn//8Z6WlpSk7O1sODg7MSwupra3V9OnTNWDAAF199dWSfvnunJ2d5ePjY1X37Pumvvuq7tyF6pSUlKiiokJubm7NcUltRn1zI0m33HKLwsPDFRISop07d+rBBx9UQUGB3n33XUnMTXP66quvlJiYqMrKSnl4eGjt2rXq1q2b8vPzuWcagcQGAAAA7NrNN99s+TkuLk49evRQ586dtXXrVo0YMcKGkV1Zpk6dqq+//lqffvqprUPBWc43N7/dZyYuLk7BwcEaMWKECgsL1blz55YO84oSExOj/Px8nTx5Um+//bYyMjKUlZVl67DsFktRAAAAGiAoKEiSztmZ/tChQ5ZzQUFBOnz4sNX56upqHTt2zKpOfX38dgxcnsjISPn7+2vPnj2SmJeWMG3aNK1fv16ffPKJOnbsaCkPCgpSVVWVTpw4YVX/7PvmYt/9+ep4eXm1uX95bmrnm5v69OvXT5Ks7h3mpnk4OzsrKipKffr00YIFCxQfH69nnnmGe6aRSGwAAAA0QEREhIKCgpSZmWkpKykpUU5OjhITEyVJiYmJOnHihPLy8ix1tmzZotraWssfDImJidq2bZvOnDljqbNp0ybFxMSw3KGJ/Pvf/9bRo0cVHBwsiXlpToZhaNq0aVq7dq22bNlyznKePn36yMnJyeq+KSgoUHFxsdV989VXX1klnzZt2iQvLy9169bNUue3fdTVqesD57rY3NQnPz9fkqzuHeamZdTW1ur06dPcM41l691LAQAAWovS0lJjx44dxo4dOwxJxuLFi40dO3YY+/btMwzDMJ566inDx8fHeO+994ydO3cao0ePNiIiIoyKigpLH6mpqUavXr2MnJwc49NPPzWio6ON8ePHW86fOHHCCAwMNG699Vbj66+/Nt544w3D3d3deOmll1r8eu3FhealtLTUmDVrlpGdnW0UFRUZmzdvNnr37m1ER0cblZWVlj6Yl+Zx1113Gd7e3sbWrVuNAwcOWD6nTp2y1LnzzjuNsLAwY8uWLUZubq6RmJhoJCYmWs5XV1cbV199tTFy5EgjPz/f2Lhxo9GhQwfj4YcfttT54YcfDHd3d+P+++83du/ebSxdutRwcHAwNm7c2KLXa08uNjd79uwxHn/8cSM3N9coKioy3nvvPSMyMtIYPHiwpQ/mpnk89NBDRlZWllFUVGTs3LnTeOihhwyTyWR8/PHHhmFwzzQGiQ0AAIBfffLJJ4akcz4ZGRmGYfzyytdHH33UCAwMNFxcXIwRI0YYBQUFVn0cPXrUGD9+vOHh4WF4eXkZt99+u1FaWmpV58svvzQGDhxouLi4GFdddZXx1FNPtdQl2qULzcupU6eMkSNHGh06dDCcnJyM8PBwY8qUKVavQTQM5qW51DcvkowVK1ZY6lRUVBh333234evra7i7uxs33HCDceDAAat+9u7da6SlpRlubm6Gv7+/MXPmTOPMmTNWdT755BOjZ8+ehrOzsxEZGWk1Bs51sbkpLi42Bg8ebPj5+RkuLi5GVFSUcf/99xsnT5606oe5aXp//OMfjfDwcMPZ2dno0KGDMWLECEtSwzC4ZxrDZBiG0XLPhwAAAAAAADQd9tgAAAAAAAB2i8QGAAAAAACwWyQ2AAAAAACA3SKxAQAAAAAA7BaJDQAAAAAAYLdIbAAAAAAAALtFYgMAAAAAANgtEhsAAAAAAMBukdgAAAAAcMXo1KmTlixZcsE6VVVVioqK0j//+c8G97tx40b17NlTtbW1lxkhgEtFYgMAAAC4RBMnTtSYMWNsHQaaybJlyxQREaH+/ftblWdmZmrYsGHy8/OTj4+PbrvtNv3000+SpNTUVDk5Oen111+3RcjAFY3EBgAAANBKnTlzxtYhXHEMw9Dzzz+vSZMmWZWvX79eaWlpGjFihD777DNlZWWpZ8+eWrNmjaXOxIkT9eyzz7Z0yMAVj8QGAAAAcJmGDh2qe++9Vw888ID8/PwUFBSkefPmWdX59ttvNXDgQLm6uqpbt27avHmzTCaT1q1bJ0nau3evTCaT1qxZoyFDhsjV1dXyr/+vvPKKYmNj5erqqq5du+qFF16w9FtVVaVp06YpODhYrq6uCg8P14IFCyzni4uLNXr0aHl4eMjLy0s33XSTDh06dN5rqYvjzTff1KBBg+Tm5qaEhAR99913+uKLL9S3b195eHgoLS1NR44csWp7oTjr+n333Xc1bNgwubu7Kz4+XtnZ2ZY6+/btU3p6unx9fdWuXTt1795dH374oSRp5cqV8vHxsRpv3bp1MplMluPCwkKNHj1agYGB8vDwUEJCgjZv3nyBmTtXXl6eCgsLNWrUKKvyOXPmaNq0aXrkkUcUGxur+Ph4zZgxQ9OnT7fUSU9PV25urgoLCy9pTACXx9HWAQAAAABtwauvvqoZM2YoJydH2dnZmjhxogYMGKDk5GTV1NRozJgxCgsLU05OjkpLSzVz5sx6+3nooYe0aNEi9erVy5LcmDNnjp5//nn16tVLO3bs0JQpU9SuXTtlZGTo2Wef1fvvv68333xTYWFh2r9/v/bv3y9Jqq2ttSQ1srKyVF1dralTp2rcuHHaunXrBa9n7ty5WrJkicLCwvTHP/5Rt9xyizw9PfXMM8/I3d1dN910k+bMmaMXX3xRki4aZ53Zs2frr3/9q6KjozV79myNHz9ee/bskaOjo6ZOnaqqqipt27ZN7dq1065du+Th4dHgOSgrK9N1112nJ598Ui4uLlq1apXS09NVUFCgsLCwBvWxfft2denSRZ6enpKknJwcjRgxQuXl5dq1a5defvllSZLJZNLXX3+t8PBwS9uwsDAFBgZq+/bt6ty5c4PjBnB5SGwAAAAATaBHjx6aO3euJCk6OlrPP/+8MjMzlZycrE2bNqmwsFBbt25VUFCQJOnJJ59UcnLyOf1Mnz5dN954o+V47ty5WrRokaUsIiJCu3bt0ksvvaSMjAwVFxcrOjpaAwcOlMlksvpDOzMzU1999ZWKiooUGhoqSVq1apW6d++uL774QgkJCee9nlmzZiklJUWSdN9992n8+PHKzMzUgAEDJEmTJk3SypUrGxznb/utexriscceU/fu3bVnzx517dpVxcXFGjt2rOLi4iRJkZGRDfnqLeLj4xUfH285fuKJJ7R27Vq9//77mjZtWoP62Ldvn0JCQizHsbGx+uCDDzR06FC9+eab6tatmyTJwcHB6ruuExISon379l1S3AAuD4kNAAAAoAn06NHD6jg4OFiHDx+WJBUUFCg0NNSS1JCka665pt5++vbta/m5vLxchYWFmjRpkqZMmWIpr66ulre3t6Rf9nVITk5WTEyMUlNTdf3112vkyJGSpN27dys0NNSS1JCkbt26ycfHR7t3775gYuO31xMYGChJloRDXVnd9TUkzvr6DQ4OliQdPnxYXbt21b333qu77rpLH3/8sZKSkjR27NhzvtcLKSsr07x58/TBBx/owIEDqq6uVkVFhYqLixvcR0VFhVxdXS3Hu3fv1nXXXSdJuvnmm2U2m2U2m7Vr165627u5uenUqVMNHg/A5SOxAQAAADQBJycnq2OTydSoV3+2a9fO8nNZWZkkafny5erXr59VPQcHB0lS7969VVRUpA0bNmjz5s266aablJSUpLfffvuSx/6t315P3T4WZ5fVXV9D4rxQv3X9TJ48WSkpKfrggw/08ccfa8GCBVq0aJHuuecemc1mGYZh1dfZm6vOmjVLmzZt0l//+ldFRUXJzc1Nv//971VVVdXg6/b399dXX31lOY6NjdWXX36pIUOG6LbbbtOkSZPk6Oiojh071tv+2LFj6tChQ4PHA3D52DwUAAAAaGYxMTHav3+/1aadX3zxxUXbBQYGKiQkRD/88IOioqKsPhEREZZ6Xl5eGjdunJYvX641a9bonXfe0bFjxxQbG2u154Yk7dq1SydOnLAsqWgKDY2zIUJDQ3XnnXfq3Xff1cyZM7V8+XJJUocOHVRaWqry8nJL3fz8fKu2n332mSZOnKgbbrhBcXFxCgoK0t69ey9p/F69eunbb7+1JFG8vLwUFRWlUaNGafPmzYqMjFSnTp0s9X/73VZWVqqwsFC9evW6pDEBXB6e2AAAAACaWXJysjp37qyMjAwtXLhQpaWleuSRRyTJ6q0e9Xnsscd07733ytvbW6mpqTp9+rRyc3N1/PhxzZgxQ4sXL1ZwcLB69eols9mst956S0FBQfLx8VFSUpLi4uI0YcIELVmyRNXV1br77rs1ZMgQqyUvTeFicTbE9OnTlZaWpi5duuj48eP65JNPFBsbK0nq16+f3N3d9ec//1n33nuvcnJyrPb4kH7Z2+Tdd99Venq6TCaTHn300Ut+ambYsGEqKyvTN998o6uvvtpSPm/ePMXFxenWW2/VAw88ICcnJz3//PMKDAy07K3y+eefy8XFRYmJiZc0JoDLwxMbAAAAQDNzcHDQunXrVFZWpoSEBE2ePFmzZ8+WJKv9HOozefJkvfLKK1qxYoXi4uI0ZMgQrVy50vIkhKenpxYuXKi+ffsqISFBe/fu1Ycffiiz2SyTyaT33ntPvr6+Gjx4sJKSkhQZGak1a9Y0+TVeLM6GqKmp0dSpUxUbG6vU1FR16dLF8spYPz8/vfbaa/rwww8VFxen1atXn/NK3cWLF8vX11f9+/dXenq6UlJS1Lt370u6jvbt2+uGG26wvGq3TkhIiLKysnT8+HENGDBA/fv3tySK6qxevVoTJkyQu7v7JY0J4PKYjLMXqgEAAABodp999pkGDhyoPXv28GrQVmbnzp1KTk5WYWFhg183+/PPPysmJka5ubmXvPwGwOUhsQEAAAC0gLVr18rDw0PR0dHas2eP7rvvPvn6+urTTz+1dWiox8qVK9WnTx+rN8FcSG5urgoLCzVu3LhmjgzA2UhsAAAAAC1g1apVmj9/voqLi+Xv76+kpCQtWrRI7du3t3VoAGDXSGwAAAAAAAC7xeahAAAAAADAbpHYAAAAAAAAdovEBgAAAAAAsFskNgAAAAAAgN0isQEAAAAAAOwWiQ0AAAAAAGC3SGwAAAAAAAC7RWIDAAAAAADYLRIbAAAAAADAbv1/VADUqmmOREYAAAAASUVORK5CYII=\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "✅ Figura guardada: eda_distribucion.png\n"
          ]
        }
      ],
      "source": [
        "# Distribución de la variable objetivo\n",
        "fig, axes = plt.subplots(1, 2, figsize=(12, 4))\n",
        "\n",
        "# Pie chart\n",
        "counts = df['impago'].value_counts()\n",
        "axes[0].pie(counts, labels=['Paga (0)', 'Impago (1)'],\n",
        "            colors=['#2E86AB', '#E84855'], autopct='%1.1f%%',\n",
        "            startangle=90, textprops={'fontsize': 12})\n",
        "axes[0].set_title('Distribución variable objetivo', fontsize=13, fontweight='bold')\n",
        "\n",
        "# Histograma ingreso por clase\n",
        "for label, color in [(0, '#2E86AB'), (1, '#E84855')]:\n",
        "    axes[1].hist(df[df['impago']==label]['ingreso_mensual_eur'],\n",
        "                 bins=30, alpha=0.6, color=color,\n",
        "                 label='Paga' if label==0 else 'Impago')\n",
        "axes[1].set_xlabel('Ingreso mensual (€)')\n",
        "axes[1].set_ylabel('Frecuencia')\n",
        "axes[1].set_title('Distribución ingresos por clase', fontsize=13, fontweight='bold')\n",
        "axes[1].legend()\n",
        "\n",
        "plt.suptitle('RemiCash PoC — Análisis Exploratorio del Dataset', fontsize=14, y=1.02)\n",
        "plt.tight_layout()\n",
        "plt.savefig('eda_distribucion.png', dpi=150, bbox_inches='tight')\n",
        "plt.show()\n",
        "print('✅ Figura guardada: eda_distribucion.png')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "smVRZIeCSc4P"
      },
      "source": [
        "## 3. Preprocesamiento & Feature Engineering"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JhiEnbhcSc4P",
        "outputId": "a7af31a1-1110-44e4-e108-8a8eff24b09a"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train: 800 muestras | Test: 200 muestras\n",
            "Tasa impago train: 31.2% | test: 31.5%\n"
          ]
        }
      ],
      "source": [
        "FEATURES = [\n",
        "    'ingreso_mensual_eur',\n",
        "    'antiguedad_cuenta_meses',\n",
        "    'ratio_gasto_ingreso',\n",
        "    'frecuencia_remesas_anual',\n",
        "    'regularidad_ingresos',\n",
        "    'descubiertos_12m',\n",
        "    'variabilidad_ingresos_eur',\n",
        "    'num_cuentas_bancarias',\n",
        "    'cuotas_deuda_mensual_eur',\n",
        "    'ahorro_medio_mensual_eur'\n",
        "]\n",
        "TARGET = 'impago'\n",
        "\n",
        "X = df[FEATURES]\n",
        "y = df[TARGET]\n",
        "\n",
        "# Train / test split estratificado\n",
        "X_train, X_test, y_train, y_test = train_test_split(\n",
        "    X, y, test_size=0.2, random_state=SEED, stratify=y\n",
        ")\n",
        "\n",
        "print(f'Train: {X_train.shape[0]} muestras | Test: {X_test.shape[0]} muestras')\n",
        "print(f'Tasa impago train: {y_train.mean():.1%} | test: {y_test.mean():.1%}')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Oi-6qDudSc4P"
      },
      "source": [
        "## 4. Modelos ML — Entrenamiento y Evaluación Comparativa"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 227
        },
        "id": "BaB7GQbCSc4P",
        "outputId": "f1b43734-e856-4219-ee57-a3b845e4e47f"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Logistic Regression (baseline)           | AUC: 0.924 | F1: 0.911\n",
            "Random Forest                            | AUC: 0.925 | F1: 0.911\n",
            "XGBoost                                  | AUC: 0.934 | F1: 0.911\n",
            "\n",
            "--- Tabla comparativa ---\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                           Modelo  AUC-ROC (test)  AUC-ROC (CV-5)  F1-Score  \\\n",
              "0  Logistic Regression (baseline)          0.9243          0.9414    0.9106   \n",
              "1                   Random Forest          0.9248          0.9445    0.9106   \n",
              "2                         XGBoost          0.9337          0.9488    0.9106   \n",
              "\n",
              "   Precision  Recall  \n",
              "0     0.9333  0.8889  \n",
              "1     0.9333  0.8889  \n",
              "2     0.9333  0.8889  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-33c42655-cfbc-4fc5-82d4-bb488978f71d\" 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>Modelo</th>\n",
              "      <th>AUC-ROC (test)</th>\n",
              "      <th>AUC-ROC (CV-5)</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Logistic Regression (baseline)</td>\n",
              "      <td>0.9243</td>\n",
              "      <td>0.9414</td>\n",
              "      <td>0.9106</td>\n",
              "      <td>0.9333</td>\n",
              "      <td>0.8889</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Random Forest</td>\n",
              "      <td>0.9248</td>\n",
              "      <td>0.9445</td>\n",
              "      <td>0.9106</td>\n",
              "      <td>0.9333</td>\n",
              "      <td>0.8889</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>XGBoost</td>\n",
              "      <td>0.9337</td>\n",
              "      <td>0.9488</td>\n",
              "      <td>0.9106</td>\n",
              "      <td>0.9333</td>\n",
              "      <td>0.8889</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-33c42655-cfbc-4fc5-82d4-bb488978f71d')\"\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-33c42655-cfbc-4fc5-82d4-bb488978f71d 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-33c42655-cfbc-4fc5-82d4-bb488978f71d');\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=\"id_be0d13b9-ef9e-4aa4-acf0-b7835f6aeaf1\">\n",
              "    <style>\n",
              "      .colab-df-generate {\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-generate: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",
              "      [theme=dark] .colab-df-generate {\n",
              "        background-color: #3B4455;\n",
              "        fill: #D2E3FC;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate: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",
              "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('df_resultados')\"\n",
              "            title=\"Generate code using this dataframe.\"\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",
              "    <path d=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "    <script>\n",
              "      (() => {\n",
              "      const buttonEl =\n",
              "        document.querySelector('#id_be0d13b9-ef9e-4aa4-acf0-b7835f6aeaf1 button.colab-df-generate');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      buttonEl.onclick = () => {\n",
              "        google.colab.notebook.generateWithVariable('df_resultados');\n",
              "      }\n",
              "      })();\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "df_resultados",
              "summary": "{\n  \"name\": \"df_resultados\",\n  \"rows\": 3,\n  \"fields\": [\n    {\n      \"column\": \"Modelo\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 3,\n        \"samples\": [\n          \"Logistic Regression (baseline)\",\n          \"Random Forest\",\n          \"XGBoost\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"AUC-ROC (test)\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.005288667128870934,\n        \"min\": 0.9243,\n        \"max\": 0.9337,\n        \"num_unique_values\": 3,\n        \"samples\": [\n          0.9243,\n          0.9248,\n          0.9337\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"AUC-ROC (CV-5)\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.00371618083521239,\n        \"min\": 0.9414,\n        \"max\": 0.9488,\n        \"num_unique_values\": 3,\n        \"samples\": [\n          0.9414,\n          0.9445,\n          0.9488\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"F1-Score\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.0,\n        \"min\": 0.9106,\n        \"max\": 0.9106,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          0.9106\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Precision\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.0,\n        \"min\": 0.9333,\n        \"max\": 0.9333,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          0.9333\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Recall\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.0,\n        \"min\": 0.8889,\n        \"max\": 0.8889,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          0.8889\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 6
        }
      ],
      "source": [
        "modelos = {\n",
        "    'Logistic Regression (baseline)': LogisticRegression(\n",
        "        max_iter=1000, random_state=SEED, class_weight='balanced'\n",
        "    ),\n",
        "    'Random Forest': RandomForestClassifier(\n",
        "        n_estimators=100, random_state=SEED, class_weight='balanced'\n",
        "    ),\n",
        "    'XGBoost': XGBClassifier(\n",
        "        n_estimators=100, learning_rate=0.1, max_depth=4,\n",
        "        use_label_encoder=False, eval_metric='logloss',\n",
        "        random_state=SEED, scale_pos_weight=(y_train==0).sum()/(y_train==1).sum()\n",
        "    )\n",
        "}\n",
        "\n",
        "resultados = []\n",
        "cv = StratifiedKFold(n_splits=5, shuffle=True, random_state=SEED)\n",
        "\n",
        "for nombre, modelo in modelos.items():\n",
        "    modelo.fit(X_train, y_train)\n",
        "    y_pred = modelo.predict(X_test)\n",
        "    y_proba = modelo.predict_proba(X_test)[:, 1]\n",
        "\n",
        "    auc    = roc_auc_score(y_test, y_proba)\n",
        "    f1     = f1_score(y_test, y_pred)\n",
        "    prec   = precision_score(y_test, y_pred)\n",
        "    rec    = recall_score(y_test, y_pred)\n",
        "    cv_auc = cross_val_score(modelo, X, y, cv=cv, scoring='roc_auc').mean()\n",
        "\n",
        "    resultados.append({\n",
        "        'Modelo': nombre,\n",
        "        'AUC-ROC (test)': round(auc, 4),\n",
        "        'AUC-ROC (CV-5)': round(cv_auc, 4),\n",
        "        'F1-Score':       round(f1, 4),\n",
        "        'Precision':      round(prec, 4),\n",
        "        'Recall':         round(rec, 4)\n",
        "    })\n",
        "    print(f'{nombre:40s} | AUC: {auc:.3f} | F1: {f1:.3f}')\n",
        "\n",
        "df_resultados = pd.DataFrame(resultados)\n",
        "print('\\n--- Tabla comparativa ---')\n",
        "df_resultados"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 488
        },
        "id": "__0tGbPvSc4Q",
        "outputId": "2a417448-78b8-4c06-8f2d-c6fd41d2632f"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1400x500 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "✅ Figura guardada: metricas_modelos.png\n"
          ]
        }
      ],
      "source": [
        "# Curvas ROC comparativas\n",
        "fig, axes = plt.subplots(1, 2, figsize=(14, 5))\n",
        "colores = ['#7C83FD', '#2E86AB', '#E84855']\n",
        "\n",
        "for (nombre, modelo), color in zip(modelos.items(), colores):\n",
        "    y_proba = modelo.predict_proba(X_test)[:, 1]\n",
        "    fpr, tpr, _ = roc_curve(y_test, y_proba)\n",
        "    auc = roc_auc_score(y_test, y_proba)\n",
        "    axes[0].plot(fpr, tpr, color=color, lw=2, label=f'{nombre} (AUC={auc:.3f})')\n",
        "\n",
        "axes[0].plot([0,1],[0,1], 'k--', alpha=0.4)\n",
        "axes[0].axhline(y=0.75, color='gray', linestyle=':', alpha=0.7, label='Objetivo mínimo AUC=0.75')\n",
        "axes[0].set_xlabel('Tasa de Falsos Positivos')\n",
        "axes[0].set_ylabel('Tasa de Verdaderos Positivos')\n",
        "axes[0].set_title('Curvas ROC — Comparativa de Modelos', fontweight='bold')\n",
        "axes[0].legend(fontsize=9)\n",
        "axes[0].grid(alpha=0.3)\n",
        "\n",
        "# Bar chart métricas XGBoost (modelo seleccionado)\n",
        "metricas_xgb = df_resultados[df_resultados['Modelo']=='XGBoost'][['AUC-ROC (test)','F1-Score','Precision','Recall']].values[0]\n",
        "nombres_m = ['AUC-ROC', 'F1-Score', 'Precision', 'Recall']\n",
        "objetivos = [0.75, 0.70, 0.70, 0.70]\n",
        "bars = axes[1].bar(nombres_m, metricas_xgb, color='#2E86AB', alpha=0.8)\n",
        "axes[1].bar(nombres_m, objetivos, color='#E84855', alpha=0.3, label='Objetivo mínimo')\n",
        "for bar, val in zip(bars, metricas_xgb):\n",
        "    axes[1].text(bar.get_x() + bar.get_width()/2, bar.get_height()+0.01,\n",
        "                 f'{val:.3f}', ha='center', fontsize=10, fontweight='bold')\n",
        "axes[1].set_ylim(0, 1.1)\n",
        "axes[1].set_title('XGBoost — Métricas vs Objetivos PoC', fontweight='bold')\n",
        "axes[1].legend()\n",
        "axes[1].grid(axis='y', alpha=0.3)\n",
        "\n",
        "plt.suptitle('RemiCash PoC — Evaluación del Modelo', fontsize=13)\n",
        "plt.tight_layout()\n",
        "plt.savefig('metricas_modelos.png', dpi=150, bbox_inches='tight')\n",
        "plt.show()\n",
        "print('✅ Figura guardada: metricas_modelos.png')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 509
        },
        "id": "QFe8QvRnSc4Q",
        "outputId": "d970a546-fe96-4384-d8c6-1ef95d5c5f70"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 500x400 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "📊 Análisis de mora simulada:\n",
            "   Créditos aprobados:  140\n",
            "   Impagos reales:      7\n",
            "   Tasa de mora sim.:   5.0% (benchmark RemiCash: <3.5%)\n",
            "   ✅ Objetivo cumplido: NO\n"
          ]
        }
      ],
      "source": [
        "# Modelo seleccionado: XGBoost\n",
        "modelo_final = modelos['XGBoost']\n",
        "y_pred_final = modelo_final.predict(X_test)\n",
        "y_proba_final = modelo_final.predict_proba(X_test)[:, 1]\n",
        "\n",
        "# Matriz de confusión\n",
        "cm = confusion_matrix(y_test, y_pred_final)\n",
        "fig, ax = plt.subplots(figsize=(5, 4))\n",
        "sns.heatmap(cm, annot=True, fmt='d', cmap='Blues',\n",
        "            xticklabels=['Pred: Paga', 'Pred: Impago'],\n",
        "            yticklabels=['Real: Paga', 'Real: Impago'])\n",
        "ax.set_title('Matriz de Confusión — XGBoost', fontweight='bold')\n",
        "plt.tight_layout()\n",
        "plt.savefig('confusion_matrix.png', dpi=150, bbox_inches='tight')\n",
        "plt.show()\n",
        "\n",
        "# Tasa de mora simulada\n",
        "aprobados = (y_pred_final == 0).sum()\n",
        "mora_real = ((y_pred_final == 0) & (y_test == 1)).sum()\n",
        "tasa_mora = mora_real / aprobados if aprobados > 0 else 0\n",
        "print(f'\\n📊 Análisis de mora simulada:')\n",
        "print(f'   Créditos aprobados:  {aprobados}')\n",
        "print(f'   Impagos reales:      {mora_real}')\n",
        "print(f'   Tasa de mora sim.:   {tasa_mora:.1%} (benchmark RemiCash: <3.5%)')\n",
        "print(f'   ✅ Objetivo cumplido: {\"SÍ\" if tasa_mora < 0.035 else \"NO\"}')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "E-WOiiJ5Sc4Q"
      },
      "source": [
        "## 5. Explicabilidad XAI — Análisis SHAP"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 574
        },
        "id": "egDSYlrsSc4Q",
        "outputId": "64a93546-2b38-4bdc-d640-c43a3915e77e"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x550 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "✅ Figura SHAP guardada\n"
          ]
        }
      ],
      "source": [
        "# SHAP explainer para XGBoost\n",
        "explainer = shap.TreeExplainer(modelo_final)\n",
        "shap_values = explainer.shap_values(X_test)\n",
        "\n",
        "# SHAP summary plot\n",
        "plt.figure(figsize=(10, 6))\n",
        "shap.summary_plot(shap_values, X_test, feature_names=FEATURES,\n",
        "                  plot_type='bar', show=False, color='#2E86AB')\n",
        "plt.title('SHAP — Importancia de Variables (Top Features)', fontsize=13, fontweight='bold')\n",
        "plt.tight_layout()\n",
        "plt.savefig('shap_importancia.png', dpi=150, bbox_inches='tight')\n",
        "plt.show()\n",
        "print('✅ Figura SHAP guardada')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JenVABnHSc4Q",
        "outputId": "270c37f6-3118-4c39-f787-014a576de484"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "🔍 Top 5 variables más importantes (SHAP):\n",
            "   1. ratio_gasto_ingreso                 SHAP: 2.7053\n",
            "   2. cuotas_deuda_mensual_eur            SHAP: 0.4703\n",
            "   3. antiguedad_cuenta_meses             SHAP: 0.4143\n",
            "   4. ingreso_mensual_eur                 SHAP: 0.3147\n",
            "   5. regularidad_ingresos                SHAP: 0.3010\n"
          ]
        }
      ],
      "source": [
        "# Top 5 features por importancia SHAP\n",
        "mean_shap = np.abs(shap_values).mean(axis=0)\n",
        "top5_idx = np.argsort(mean_shap)[::-1][:5]\n",
        "top5_features = [(FEATURES[i], mean_shap[i]) for i in top5_idx]\n",
        "\n",
        "print('🔍 Top 5 variables más importantes (SHAP):')\n",
        "for i, (feat, val) in enumerate(top5_features, 1):\n",
        "    print(f'   {i}. {feat:<35} SHAP: {val:.4f}')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VNZZFHXfSc4Q"
      },
      "source": [
        "## 6. Agente LLM — Justificación de Decisión (Claude API)\n",
        "\n",
        "El agente recibe el score + SHAP values del perfil y genera una explicación en lenguaje natural: aprobado/rechazado + motivo. Cumple requisito de explicabilidad del AI Act."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4BTbtK9aSc4Q",
        "outputId": "e4f3bdd5-12ad-48a0-c0fe-52ee3b8af28e"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "✅ Funciones del agente LLM definidas\n"
          ]
        }
      ],
      "source": [
        "def construir_perfil_usuario(idx, X_test, y_proba, y_pred, shap_vals, top_features):\n",
        "    \"\"\"\n",
        "    Construye un dict con el perfil del usuario para enviarlo al agente LLM.\n",
        "    \"\"\"\n",
        "    perfil = X_test.iloc[idx].to_dict()\n",
        "    score = float(y_proba[idx])\n",
        "    decision = 'RECHAZADO' if y_pred[idx] == 1 else 'APROBADO'\n",
        "\n",
        "    # SHAP locales para este perfil\n",
        "    shap_locales = {\n",
        "        feat: float(shap_vals[idx][FEATURES.index(feat)])\n",
        "        for feat, _ in top_features\n",
        "    }\n",
        "\n",
        "    return {\n",
        "        'decision': decision,\n",
        "        'probabilidad_impago': round(score, 4),\n",
        "        'perfil_financiero': {k: round(v, 2) if isinstance(v, float) else v\n",
        "                               for k, v in perfil.items()},\n",
        "        'factores_decision_shap': shap_locales\n",
        "    }\n",
        "\n",
        "\n",
        "def agente_scoring_llm(perfil_dict, api_key):\n",
        "    \"\"\"\n",
        "    Agente LLM que genera explicación de la decisión crediticia.\n",
        "    Usa Claude API (claude-sonnet-4-20250514).\n",
        "    \"\"\"\n",
        "    client = anthropic.Anthropic(api_key=api_key)\n",
        "\n",
        "    system_prompt = \"\"\"Eres el agente de scoring crediticio de RemiCash, una plataforma fintech\n",
        "que ofrece microcréditos vinculados a remesas para migrantes latinoamericanos en España.\n",
        "\n",
        "Tu rol es explicar en lenguaje claro y empático la decisión crediticia generada por el modelo ML,\n",
        "siguiendo los requisitos de explicabilidad del AI Act (Art. 13 y 86).\n",
        "\n",
        "REGLAS:\n",
        "- Máximo 150 palabras\n",
        "- Tono empático y profesional\n",
        "- Menciona exactamente los 3 factores más relevantes (positivos o negativos)\n",
        "- Si es RECHAZADO, ofrece siempre 2 acciones concretas que el usuario puede tomar\n",
        "- Si es APROBADO, confirma el importe y condición de la remesa\n",
        "- Nunca uses jerga técnica (no menciones SHAP, modelo, XGBoost, probabilidades)\n",
        "- Escribe en español, segunda persona (\"tú\")\"\"\"\n",
        "\n",
        "    user_message = f\"\"\"Genera la justificación de la siguiente decisión crediticia:\n",
        "\n",
        "DECISIÓN: {perfil_dict['decision']}\n",
        "Probabilidad de impago: {perfil_dict['probabilidad_impago']:.1%}\n",
        "\n",
        "Perfil financiero del usuario:\n",
        "{json.dumps(perfil_dict['perfil_financiero'], indent=2, ensure_ascii=False)}\n",
        "\n",
        "Factores que más influyeron en la decisión (valores positivos = aumentan riesgo):\n",
        "{json.dumps(perfil_dict['factores_decision_shap'], indent=2, ensure_ascii=False)}\n",
        "\n",
        "Genera la explicación para el usuario.\"\"\"\n",
        "\n",
        "    response = client.messages.create(\n",
        "        model='claude-sonnet-4-20250514',\n",
        "        max_tokens=300,\n",
        "        system=system_prompt,\n",
        "        messages=[{'role': 'user', 'content': user_message}]\n",
        "    )\n",
        "\n",
        "    return response.content[0].text\n",
        "\n",
        "\n",
        "print('✅ Funciones del agente LLM definidas')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CJUQMrizSc4R",
        "outputId": "7fa1e2b5-69c1-42e8-8e2c-2d45316d434a"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "============================================================\n",
            "✅ PERFIL APROBADO\n",
            "============================================================\n",
            "Score impago: 4.4%\n",
            "Ingreso mensual: 1524.49€\n",
            "Ratio gasto/ingreso: 42.0%\n",
            "\n",
            "[⚠️ Agrega tu API key para ver la explicación del agente LLM]\n",
            "\n",
            "============================================================\n",
            "❌ PERFIL RECHAZADO\n",
            "============================================================\n",
            "Score impago: 99.7%\n",
            "Ingreso mensual: 1077.88€\n",
            "Ratio gasto/ingreso: 70.0%\n",
            "\n",
            "[⚠️ Agrega tu API key para ver la explicación del agente LLM]\n"
          ]
        }
      ],
      "source": [
        "# ⚠️ Reemplaza con tu API key de Anthropic\n",
        "# Puedes obtenerla en: https://console.anthropic.com\n",
        "API_KEY = 'CONFIGURAR_EN_VARIABLE_DE_ENTORNO'  # Clave eliminada en la version publica\n",
        "\n",
        "# --- Demo: 2 perfiles (1 aprobado, 1 rechazado) ---\n",
        "# Buscar un perfil aprobado y uno rechazado en el test set\n",
        "idx_aprobado  = np.where(y_pred_final == 0)[0][0]\n",
        "idx_rechazado = np.where(y_pred_final == 1)[0][0]\n",
        "\n",
        "for label, idx in [('✅ PERFIL APROBADO', idx_aprobado),\n",
        "                   ('❌ PERFIL RECHAZADO', idx_rechazado)]:\n",
        "\n",
        "    perfil = construir_perfil_usuario(\n",
        "        idx, X_test, y_proba_final, y_pred_final, shap_values, top5_features\n",
        "    )\n",
        "\n",
        "    print(f'\\n{\"=\"*60}')\n",
        "    print(f'{label}')\n",
        "    print(f'{\"=\"*60}')\n",
        "    print(f'Score impago: {perfil[\"probabilidad_impago\"]:.1%}')\n",
        "    print(f'Ingreso mensual: {perfil[\"perfil_financiero\"][\"ingreso_mensual_eur\"]}€')\n",
        "    print(f'Ratio gasto/ingreso: {perfil[\"perfil_financiero\"][\"ratio_gasto_ingreso\"]:.1%}')\n",
        "\n",
        "    if API_KEY and API_KEY != 'CONFIGURAR_EN_VARIABLE_DE_ENTORNO':\n",
        "        print('\\n💬 Explicación del Agente LLM:')\n",
        "        explicacion = agente_scoring_llm(perfil, API_KEY)\n",
        "        print(explicacion)\n",
        "    else:\n",
        "        print('\\n[⚠️ Agrega tu API key para ver la explicación del agente LLM]')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "c5yvhvu-Sc4R"
      },
      "source": [
        "## 7. Análisis de Errores & Riesgos"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 425
        },
        "id": "4Ik-v7a0Sc4R",
        "outputId": "aeb49eb4-16d0-4d37-e24f-c929d1be3b31"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "📋 ANÁLISIS DE ERRORES — XGBoost en test set\n",
            "==================================================\n",
            "Verdaderos Negativos (paga, pred. paga):      133 ✅\n",
            "Verdaderos Positivos (impago, pred. impago):   56 ✅\n",
            "Falsos Positivos (paga, pred. impago):          4 ⚠️  (excluye buenos pagadores)\n",
            "Falsos Negativos (impago, pred. paga):          7 🚨 (riesgo financiero real)\n",
            "==================================================\n",
            "\n",
            "Cobertura de usuarios (decisión emitida):    96.5%\n",
            "Tasa mora simulada (FN/aprobados):           5.0%\n",
            "\n",
            "\n",
            "📋 MATRIZ DE RIESGOS:\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "    ID                                           Riesgo  Impacto Probabilidad  \\\n",
              "0  R01                        Sesgo en datos sintéticos     Alto        Media   \n",
              "1  R02                   Overfitting del modelo XGBoost     Alto         Baja   \n",
              "2  R03                     Hallucination del agente LLM     Alto        Media   \n",
              "3  R04                       Incumplimiento GDPR / PSD2  Crítico         Baja   \n",
              "4  R05  Alta tasa de FP (exclusión de buenos pagadores)    Medio        Media   \n",
              "\n",
              "                                          Mitigación  \n",
              "0  Validar con datos reales anonimizados en produ...  \n",
              "1  Cross-val 5-fold implementada; regularización ...  \n",
              "2  Prompting estructurado con reglas rígidas; val...  \n",
              "3  Solo datos sintéticos en PoC; consentimiento e...  \n",
              "4  Ajustar umbral de decisión; añadir variable de...  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-55e388d1-2e5d-4a94-b5db-6b5ff324d8f8\" 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>ID</th>\n",
              "      <th>Riesgo</th>\n",
              "      <th>Impacto</th>\n",
              "      <th>Probabilidad</th>\n",
              "      <th>Mitigación</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>R01</td>\n",
              "      <td>Sesgo en datos sintéticos</td>\n",
              "      <td>Alto</td>\n",
              "      <td>Media</td>\n",
              "      <td>Validar con datos reales anonimizados en produ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>R02</td>\n",
              "      <td>Overfitting del modelo XGBoost</td>\n",
              "      <td>Alto</td>\n",
              "      <td>Baja</td>\n",
              "      <td>Cross-val 5-fold implementada; regularización ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>R03</td>\n",
              "      <td>Hallucination del agente LLM</td>\n",
              "      <td>Alto</td>\n",
              "      <td>Media</td>\n",
              "      <td>Prompting estructurado con reglas rígidas; val...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>R04</td>\n",
              "      <td>Incumplimiento GDPR / PSD2</td>\n",
              "      <td>Crítico</td>\n",
              "      <td>Baja</td>\n",
              "      <td>Solo datos sintéticos en PoC; consentimiento e...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>R05</td>\n",
              "      <td>Alta tasa de FP (exclusión de buenos pagadores)</td>\n",
              "      <td>Medio</td>\n",
              "      <td>Media</td>\n",
              "      <td>Ajustar umbral de decisión; añadir variable de...</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-55e388d1-2e5d-4a94-b5db-6b5ff324d8f8')\"\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-55e388d1-2e5d-4a94-b5db-6b5ff324d8f8 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-55e388d1-2e5d-4a94-b5db-6b5ff324d8f8');\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=\"id_fac8f6cb-e23d-4ed3-b9f2-b7c38d17679a\">\n",
              "    <style>\n",
              "      .colab-df-generate {\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-generate: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",
              "      [theme=dark] .colab-df-generate {\n",
              "        background-color: #3B4455;\n",
              "        fill: #D2E3FC;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate: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",
              "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('df_riesgos')\"\n",
              "            title=\"Generate code using this dataframe.\"\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",
              "    <path d=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "    <script>\n",
              "      (() => {\n",
              "      const buttonEl =\n",
              "        document.querySelector('#id_fac8f6cb-e23d-4ed3-b9f2-b7c38d17679a button.colab-df-generate');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      buttonEl.onclick = () => {\n",
              "        google.colab.notebook.generateWithVariable('df_riesgos');\n",
              "      }\n",
              "      })();\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "df_riesgos",
              "summary": "{\n  \"name\": \"df_riesgos\",\n  \"rows\": 5,\n  \"fields\": [\n    {\n      \"column\": \"ID\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 5,\n        \"samples\": [\n          \"R02\",\n          \"R05\",\n          \"R03\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Riesgo\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 5,\n        \"samples\": [\n          \"Overfitting del modelo XGBoost\",\n          \"Alta tasa de FP (exclusi\\u00f3n de buenos pagadores)\",\n          \"Hallucination del agente LLM\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Impacto\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 3,\n        \"samples\": [\n          \"Alto\",\n          \"Cr\\u00edtico\",\n          \"Medio\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Probabilidad\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          \"Baja\",\n          \"Media\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Mitigaci\\u00f3n\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 5,\n        \"samples\": [\n          \"Cross-val 5-fold implementada; regularizaci\\u00f3n L2; monitoreo AUC en prod\",\n          \"Ajustar umbral de decisi\\u00f3n; a\\u00f1adir variable de comportamiento de remesas\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 13
        }
      ],
      "source": [
        "# Análisis de errores por tipo\n",
        "falsos_positivos = ((y_pred_final == 1) & (y_test.values == 0)).sum()\n",
        "falsos_negativos = ((y_pred_final == 0) & (y_test.values == 1)).sum()\n",
        "verdaderos_pos   = ((y_pred_final == 1) & (y_test.values == 1)).sum()\n",
        "verdaderos_neg   = ((y_pred_final == 0) & (y_test.values == 0)).sum()\n",
        "\n",
        "print('📋 ANÁLISIS DE ERRORES — XGBoost en test set')\n",
        "print(f'{'='*50}')\n",
        "print(f'Verdaderos Negativos (paga, pred. paga):     {verdaderos_neg:4d} ✅')\n",
        "print(f'Verdaderos Positivos (impago, pred. impago): {verdaderos_pos:4d} ✅')\n",
        "print(f'Falsos Positivos (paga, pred. impago):       {falsos_positivos:4d} ⚠️  (excluye buenos pagadores)')\n",
        "print(f'Falsos Negativos (impago, pred. paga):       {falsos_negativos:4d} 🚨 (riesgo financiero real)')\n",
        "print(f'{'='*50}')\n",
        "print(f'\\nCobertura de usuarios (decisión emitida):    {(1 - (falsos_negativos/len(y_test))):.1%}')\n",
        "print(f'Tasa mora simulada (FN/aprobados):           {tasa_mora:.1%}')\n",
        "\n",
        "riesgos = [\n",
        "    {\n",
        "        'ID': 'R01',\n",
        "        'Riesgo': 'Sesgo en datos sintéticos',\n",
        "        'Impacto': 'Alto',\n",
        "        'Probabilidad': 'Media',\n",
        "        'Mitigación': 'Validar con datos reales anonimizados en producción; test de fairness por origen'\n",
        "    },\n",
        "    {\n",
        "        'ID': 'R02',\n",
        "        'Riesgo': 'Overfitting del modelo XGBoost',\n",
        "        'Impacto': 'Alto',\n",
        "        'Probabilidad': 'Baja',\n",
        "        'Mitigación': 'Cross-val 5-fold implementada; regularización L2; monitoreo AUC en prod'\n",
        "    },\n",
        "    {\n",
        "        'ID': 'R03',\n",
        "        'Riesgo': 'Hallucination del agente LLM',\n",
        "        'Impacto': 'Alto',\n",
        "        'Probabilidad': 'Media',\n",
        "        'Mitigación': 'Prompting estructurado con reglas rígidas; validación de output antes de mostrar al usuario'\n",
        "    },\n",
        "    {\n",
        "        'ID': 'R04',\n",
        "        'Riesgo': 'Incumplimiento GDPR / PSD2',\n",
        "        'Impacto': 'Crítico',\n",
        "        'Probabilidad': 'Baja',\n",
        "        'Mitigación': 'Solo datos sintéticos en PoC; consentimiento explícito en arquitectura productiva'\n",
        "    },\n",
        "    {\n",
        "        'ID': 'R05',\n",
        "        'Riesgo': 'Alta tasa de FP (exclusión de buenos pagadores)',\n",
        "        'Impacto': 'Medio',\n",
        "        'Probabilidad': 'Media',\n",
        "        'Mitigación': 'Ajustar umbral de decisión; añadir variable de comportamiento de remesas'\n",
        "    }\n",
        "]\n",
        "\n",
        "df_riesgos = pd.DataFrame(riesgos)\n",
        "print('\\n\\n📋 MATRIZ DE RIESGOS:')\n",
        "df_riesgos"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7zzn6TjWSc4R"
      },
      "source": [
        "## 8. Resumen de Resultados"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rCyUrXEJSc4R",
        "outputId": "d3f778fc-53cd-416b-cc3b-4b59e2011551"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "============================================================\n",
            "📊 RESUMEN DE RESULTADOS — RemiCash PoC Entrega 2\n",
            "============================================================\n",
            "Modelo seleccionado: XGBoost\n",
            "Dataset: 1000 perfiles sintéticos (Open Banking simulado)\n",
            "\n",
            "MÉTRICAS:\n",
            "  AUC-ROC (test):   0.934   (objetivo ≥0.75 ✅)\n",
            "  F1-Score:         0.911   (objetivo ≥0.70 ✅)\n",
            "  Tasa mora sim.:   5.0%   (objetivo <3.5% ❌)\n",
            "\n",
            "COMPONENTES IMPLEMENTADOS:\n",
            "  ✅ Dataset Open Banking sintético (1000 perfiles)\n",
            "  ✅ Modelo XGBoost + comparativa LR / RF\n",
            "  ✅ Validación cruzada k-fold (k=5)\n",
            "  ✅ Análisis SHAP (Top-5 features identificadas)\n",
            "  ✅ Agente LLM (Claude API) — decisiones explicables\n",
            "  ✅ Matriz de riesgos + plan de mitigación\n",
            "\n",
            "TOP 5 FEATURES (SHAP):\n",
            "  1. ratio_gasto_ingreso\n",
            "  2. cuotas_deuda_mensual_eur\n",
            "  3. antiguedad_cuenta_meses\n",
            "  4. ingreso_mensual_eur\n",
            "  5. regularidad_ingresos\n",
            "============================================================\n"
          ]
        }
      ],
      "source": [
        "xgb_row = df_resultados[df_resultados['Modelo']=='XGBoost'].iloc[0]\n",
        "\n",
        "print('='*60)\n",
        "print('📊 RESUMEN DE RESULTADOS — RemiCash PoC Entrega 2')\n",
        "print('='*60)\n",
        "print(f'Modelo seleccionado: XGBoost')\n",
        "print(f'Dataset: {len(df)} perfiles sintéticos (Open Banking simulado)')\n",
        "print()\n",
        "print('MÉTRICAS:')\n",
        "print(f'  AUC-ROC (test):   {xgb_row[\"AUC-ROC (test)\"]:.3f}   (objetivo ≥0.75 ✅)'\n",
        "      if xgb_row['AUC-ROC (test)'] >= 0.75 else\n",
        "      f'  AUC-ROC (test):   {xgb_row[\"AUC-ROC (test)\"]:.3f}   (objetivo ≥0.75 ❌)')\n",
        "print(f'  F1-Score:         {xgb_row[\"F1-Score\"]:.3f}   (objetivo ≥0.70 {\"✅\" if xgb_row[\"F1-Score\"]>=0.70 else \"❌\"})')\n",
        "print(f'  Tasa mora sim.:   {tasa_mora:.1%}   (objetivo <3.5% {\"✅\" if tasa_mora<0.035 else \"❌\"})')\n",
        "print()\n",
        "print('COMPONENTES IMPLEMENTADOS:')\n",
        "print('  ✅ Dataset Open Banking sintético (1000 perfiles)')\n",
        "print('  ✅ Modelo XGBoost + comparativa LR / RF')\n",
        "print('  ✅ Validación cruzada k-fold (k=5)')\n",
        "print('  ✅ Análisis SHAP (Top-5 features identificadas)')\n",
        "print('  ✅ Agente LLM (Claude API) — decisiones explicables')\n",
        "print('  ✅ Matriz de riesgos + plan de mitigación')\n",
        "print()\n",
        "print('TOP 5 FEATURES (SHAP):')\n",
        "for i, (feat, val) in enumerate(top5_features, 1):\n",
        "    print(f'  {i}. {feat}')\n",
        "print('='*60)"
      ]
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "name": "python",
      "version": "3.10.0"
    },
    "colab": {
      "provenance": []
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}