{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "8cc86561",
   "metadata": {},
   "source": [
    "\n",
    "# Поиск аномалий в сельскохозяйственных данных (2013–2015)\n",
    "\n",
    "Задача:\n",
    "1. Отобрать данные за **2013–2015 годы**\n",
    "2. Использовать признаки:\n",
    "   - **Динамика посевных площадей**\n",
    "   - **Динамика количества КРС**\n",
    "3. Найти аномалии **двумя способами**:\n",
    "   - Без нейросетей (Isolation Forest)\n",
    "   - С использованием нейросетей (Autoencoder)\n",
    "4. Рассчитать **метрики точности**\n",
    "5. Построить **графики**\n",
    "6. Сделать **вывод о лучшем методе**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "87a682c8",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.ensemble import IsolationForest\n",
    "from sklearn.metrics import precision_score, recall_score, f1_score\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "916df76a",
   "metadata": {},
   "source": [
    "## 1. Загрузка данных"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "257f97f1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>Район</th>\n",
       "      <th>Год</th>\n",
       "      <th>Динамика_Количества_Работников</th>\n",
       "      <th>Динамика_Величина_КРС</th>\n",
       "      <th>Динамика_Мощности_Техники</th>\n",
       "      <th>Динамика_Количества_Техники</th>\n",
       "      <th>Динамика_Основных_Средств</th>\n",
       "      <th>Динамика_Посевных_Площадей</th>\n",
       "      <th>Динамика_Себестоимости_Произведенной_Продукции</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Арзамасский</td>\n",
       "      <td>2019</td>\n",
       "      <td>0.9648</td>\n",
       "      <td>0.8773</td>\n",
       "      <td>1.0907</td>\n",
       "      <td>1.0638</td>\n",
       "      <td>1.3537</td>\n",
       "      <td>1.0871</td>\n",
       "      <td>1.4476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Балахнинский</td>\n",
       "      <td>2019</td>\n",
       "      <td>1.0125</td>\n",
       "      <td>0.9695</td>\n",
       "      <td>1.0091</td>\n",
       "      <td>0.9500</td>\n",
       "      <td>1.0919</td>\n",
       "      <td>1.3471</td>\n",
       "      <td>1.0623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Богородский</td>\n",
       "      <td>2019</td>\n",
       "      <td>0.8421</td>\n",
       "      <td>0.9707</td>\n",
       "      <td>1.0060</td>\n",
       "      <td>0.9636</td>\n",
       "      <td>1.0791</td>\n",
       "      <td>0.7955</td>\n",
       "      <td>1.0837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Большеболдинский</td>\n",
       "      <td>2019</td>\n",
       "      <td>0.7951</td>\n",
       "      <td>0.8774</td>\n",
       "      <td>0.8965</td>\n",
       "      <td>0.8611</td>\n",
       "      <td>0.9460</td>\n",
       "      <td>0.9489</td>\n",
       "      <td>0.7530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Борский</td>\n",
       "      <td>2019</td>\n",
       "      <td>0.8389</td>\n",
       "      <td>0.7987</td>\n",
       "      <td>0.8097</td>\n",
       "      <td>1.0094</td>\n",
       "      <td>0.8950</td>\n",
       "      <td>0.8701</td>\n",
       "      <td>1.0847</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Район   Год  Динамика_Количества_Работников  \\\n",
       "0       Арзамасский  2019                          0.9648   \n",
       "1      Балахнинский  2019                          1.0125   \n",
       "2       Богородский  2019                          0.8421   \n",
       "3  Большеболдинский  2019                          0.7951   \n",
       "4           Борский  2019                          0.8389   \n",
       "\n",
       "   Динамика_Величина_КРС  Динамика_Мощности_Техники  \\\n",
       "0                 0.8773                     1.0907   \n",
       "1                 0.9695                     1.0091   \n",
       "2                 0.9707                     1.0060   \n",
       "3                 0.8774                     0.8965   \n",
       "4                 0.7987                     0.8097   \n",
       "\n",
       "   Динамика_Количества_Техники  Динамика_Основных_Средств  \\\n",
       "0                       1.0638                     1.3537   \n",
       "1                       0.9500                     1.0919   \n",
       "2                       0.9636                     1.0791   \n",
       "3                       0.8611                     0.9460   \n",
       "4                       1.0094                     0.8950   \n",
       "\n",
       "   Динамика_Посевных_Площадей  Динамика_Себестоимости_Произведенной_Продукции  \n",
       "0                      1.0871                                          1.4476  \n",
       "1                      1.3471                                          1.0623  \n",
       "2                      0.7955                                          1.0837  \n",
       "3                      0.9489                                          0.7530  \n",
       "4                      0.8701                                          1.0847  "
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "df = pd.read_csv(\"Reproduction_Agriculture_Resources.csv\", encoding=\"cp1251\", sep=\";\")\n",
    "\n",
    "df.head()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "45734090",
   "metadata": {},
   "source": [
    "## 2. Отбор данных за 2013–2015"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "68fedc56",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<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>Динамика_Посевных_Площадей</th>\n",
       "      <th>Динамика_Величина_КРС</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>1.0339</td>\n",
       "      <td>0.9851</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>129</th>\n",
       "      <td>1.0206</td>\n",
       "      <td>1.0362</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>1.0000</td>\n",
       "      <td>0.9671</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131</th>\n",
       "      <td>0.9943</td>\n",
       "      <td>0.9623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>132</th>\n",
       "      <td>0.9407</td>\n",
       "      <td>0.9415</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Динамика_Посевных_Площадей  Динамика_Величина_КРС\n",
       "128                      1.0339                 0.9851\n",
       "129                      1.0206                 1.0362\n",
       "130                      1.0000                 0.9671\n",
       "131                      0.9943                 0.9623\n",
       "132                      0.9407                 0.9415"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "data = df[(df[\"Год\"] >= 2013) & (df[\"Год\"] <= 2015)]\n",
    "\n",
    "features = data[[\n",
    "    \"Динамика_Посевных_Площадей\",\n",
    "    \"Динамика_Величина_КРС\"\n",
    "]].copy()\n",
    "\n",
    "features.head()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "be11f495",
   "metadata": {},
   "source": [
    "## 3. Нормализация данных"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "b29ee9b1",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "scaler = StandardScaler()\n",
    "X = scaler.fit_transform(features)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8982e796",
   "metadata": {},
   "source": [
    "## 4. Создание псевдо‑разметки аномалий (IQR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "d9ed6ecd",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "Q1 = features.quantile(0.25)\n",
    "Q3 = features.quantile(0.75)\n",
    "IQR = Q3 - Q1\n",
    "\n",
    "lower = Q1 - 1.5 * IQR\n",
    "upper = Q3 + 1.5 * IQR\n",
    "\n",
    "labels = ((features < lower) | (features > upper)).any(axis=1).astype(int)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5e85108a",
   "metadata": {},
   "source": [
    "## 5. Метод 1 — Isolation Forest (без нейросетей)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "4cfc0350",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "iso = IsolationForest(contamination=0.1, random_state=42)\n",
    "pred_iso = iso.fit_predict(X)\n",
    "\n",
    "pred_iso = np.where(pred_iso == -1, 1, 0)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "538ece1b",
   "metadata": {},
   "source": [
    "### Метрики"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "d3f46879",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.9285714285714286, 0.6190476190476191, 0.7428571428571429)"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "precision_iso = precision_score(labels, pred_iso)\n",
    "recall_iso = recall_score(labels, pred_iso)\n",
    "f1_iso = f1_score(labels, pred_iso)\n",
    "\n",
    "precision_iso, recall_iso, f1_iso\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "07909769",
   "metadata": {},
   "source": [
    "## 6. Метод 2 — Нейросеть (Autoencoder)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "4ef02a20",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2026-03-13 16:28:10.863292: I tensorflow/core/platform/cpu_feature_guard.cc:210] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.\n",
      "To enable the following instructions: AVX2 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "WARNING: All log messages before absl::InitializeLog() is called are written to STDERR\n",
      "I0000 00:00:1773408493.324542   59989 gpu_device.cc:2020] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 7537 MB memory:  -> device: 0, name: NVIDIA GeForce RTX 3080, pci bus id: 0000:06:00.0, compute capability: 8.6\n",
      "2026-03-13 16:28:14.384585: I external/local_xla/xla/service/service.cc:163] XLA service 0x7f9ad8008ca0 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:\n",
      "2026-03-13 16:28:14.384599: I external/local_xla/xla/service/service.cc:171]   StreamExecutor device (0): NVIDIA GeForce RTX 3080, Compute Capability 8.6\n",
      "2026-03-13 16:28:14.399308: I tensorflow/compiler/mlir/tensorflow/utils/dump_mlir_util.cc:269] disabling MLIR crash reproducer, set env var `MLIR_CRASH_REPRODUCER_DIRECTORY` to enable.\n",
      "2026-03-13 16:28:14.502984: I external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:473] Loaded cuDNN version 90300\n",
      "I0000 00:00:1773408494.719888   60114 device_compiler.h:196] Compiled cluster using XLA!  This line is logged at most once for the lifetime of the process.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.src.callbacks.history.History at 0x7f9bf4882ae0>"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.layers import Input, Dense\n",
    "\n",
    "input_dim = X.shape[1]\n",
    "\n",
    "input_layer = Input(shape=(input_dim,))\n",
    "\n",
    "encoded = Dense(4, activation=\"relu\")(input_layer)\n",
    "encoded = Dense(2, activation=\"relu\")(encoded)\n",
    "\n",
    "decoded = Dense(4, activation=\"relu\")(encoded)\n",
    "decoded = Dense(input_dim, activation=\"linear\")(decoded)\n",
    "\n",
    "autoencoder = Model(input_layer, decoded)\n",
    "autoencoder.compile(optimizer=\"adam\", loss=\"mse\")\n",
    "\n",
    "autoencoder.fit(X, X, epochs=50, batch_size=16, verbose=0)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "92c6c4dd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m5/5\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 24ms/step\n"
     ]
    }
   ],
   "source": [
    "\n",
    "recon = autoencoder.predict(X)\n",
    "mse = np.mean(np.power(X - recon, 2), axis=1)\n",
    "\n",
    "threshold = np.percentile(mse, 90)\n",
    "\n",
    "pred_nn = (mse > threshold).astype(int)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "837fddfe",
   "metadata": {},
   "source": [
    "### Метрики нейросети"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "5e1eb78a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.9285714285714286, 0.6190476190476191, 0.7428571428571429)"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "precision_nn = precision_score(labels, pred_nn)\n",
    "recall_nn = recall_score(labels, pred_nn)\n",
    "f1_nn = f1_score(labels, pred_nn)\n",
    "\n",
    "precision_nn, recall_nn, f1_nn\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9b6df723",
   "metadata": {},
   "source": [
    "## 7. Визуализация результатов"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "0c986866",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjcAAAHHCAYAAABDUnkqAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjgsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvwVt1zgAAAAlwSFlzAAAPYQAAD2EBqD+naQAAWEBJREFUeJzt3Xl4U1X6B/BvWtKkbZrQhW5YKYvsWjbBCiOLxSqICypQRBYFF1CwzE+EkVKYoujIJkJBGQUUKKgUVwZFFhcsU1nKyIAgWBChC7Sl6UKTktzfH50E0qRt0ia5ye338zx5HnLvuTfn3pTet+e85xyZIAgCiIiIiCTCR+wKEBERETkTgxsiIiKSFAY3REREJCkMboiIiEhSGNwQERGRpDC4ISIiIklhcENERESSwuCGiIiIJIXBDREREUkKgxsiojrs27cPMpkM+/btc+p5ZTIZ5s+f79RzEtF1DG6o2UtPT4dMJkO/fv3Ergo1gkwmw/PPPy92Nazs2LHD4wKY+fPnQyaT2XytWbNG7OpZqaysxPz5850eXJL0tRC7AkRi27RpE2JjY5GdnY3Tp0+jQ4cOYleJJGDHjh1YtWqVzQDn6tWraNFCvF+/q1evhkqlstjmicF9ZWUlFixYAAAYNGiQuJUhr8Lghpq13Nxc/PTTT8jMzMQzzzyDTZs2ITU1VexqkcQplUpRP//RRx9FWFiY089bUVGBwMBAp5+XyFHslqJmbdOmTQgODsbw4cPx6KOPYtOmTVZlzp49C5lMhsWLF1vt6969u9VflIWFhXjqqacQEREBpVKJuLg4bNiwweY5ZTIZPv30U4t9VVVVCA4OtvrMc+fOYerUqejUqRP8/f0RGhqKxx57DGfPnrV5bYMGDbLZ/bB+/XqLMrXr//PPP5vL3shWnsibb74JmUzmcX9Vv/322+jWrRsCAgIQHByMPn36YPPmzRZljhw5gvvuuw9qtRoqlQp33303Dhw40OC5f/jhBzz22GO4+eaboVAoEBMTg+TkZFy9etVcZuLEiVi1ahUAWNx7E1v30p76rF+/HjKZDPv378fMmTPRqlUrBAYG4uGHH8alS5ccvU11+vjjj9G7d2/4+/sjLCwM48aNw4ULFyzKTJw4ESqVCmfOnMGwYcMQFBSExx9/HABgNBqxfPlydOvWDUqlEhEREXjmmWdQUlJicY6DBw8iMTERYWFh8Pf3R9u2bfHkk08CqPk/0qpVKwDAggULzPfQ07r6yDOx5YaatU2bNmHkyJHw8/NDUlISVq9ejZ9//hm33357o8539epVDBo0CKdPn8bzzz+Ptm3b4uOPP8bEiRNx5coVzJgxw6K8UqnEunXr8NBDD5m3ZWZmoqqqyurcP//8M3766SeMGTMGN910E86ePYvVq1dj0KBBOH78OAICAqyO6dy5M1555RUAwOXLl5GcnNzgNbz88st2XeuVK1ewaNEiu8q609q1azF9+nQ8+uijmDFjBqqqqvCf//wH//73vzF27FgAwH//+1/85S9/gVqtxqxZsyCXy/HOO+9g0KBB+O677+rtovn4449RWVmJ5557DqGhocjOzsbbb7+NP//8Ex9//DEA4JlnnsHFixexa9cufPjhhw3W2dH6vPDCCwgODkZqairOnj2L5cuX4/nnn8fWrVvtukfFxcUW7319fREcHAygJoCaNGkSbr/9dixatAgFBQV46623sH//fhw5cgQtW7Y0H3ft2jUkJiZiwIABWLx4sfln8JlnnjGfZ/r06cjNzcXKlStx5MgR7N+/H3K5HIWFhbjnnnvQqlUrzJ49Gy1btsTZs2eRmZkJAGjVqhVWr16N5557Dg8//DBGjhwJALjtttvsukZq5gSiZurgwYMCAGHXrl2CIAiC0WgUbrrpJmHGjBkW5XJzcwUAwptvvml1jm7dugkDBw40v1++fLkAQNi4caN5m16vF+Lj4wWVSiVotVqLcyYlJQktWrQQ8vPzzeXvvvtuYezYsVafWVlZafX5WVlZAgDhgw8+sNrXv39/YfDgwVbXsW7dOvO2gQMHWtR/x44dAgDh3nvvFWr/egAgpKammt/PmjVLCA8PF3r37m1xDncDIEybNs38/sEHHxS6detW7zEPPfSQ4OfnJ5w5c8a87eLFi0JQUJBw1113mbft3btXACDs3bvXvM3W97Bo0SJBJpMJ586dM2+bNm2a1T28sc433kt767Nu3ToBgJCQkCAYjUbz9uTkZMHX11e4cuVKvdedmpoqALB6tWnTRhCEmp/V8PBwoXv37sLVq1fNx3355ZcCAGHevHnmbRMmTBAACLNnz7b4jB9++EEAIGzatMli+86dOy22b9++XQAg/Pzzz3XW99KlS1b3isge7JaiZmvTpk2IiIjA4MGDAdR0FYwePRpbtmyBwWCwKl9ZWYnLly9bvGqX27FjByIjI5GUlGTeJpfLMX36dJSXl+O7776zKN+rVy9069bN/Nf9uXPnsHfvXkycONHq8/39/c3/rq6uRlFRETp06ICWLVvi8OHDVuX1ej0UCoXd90MQBMyZMwePPPJIg8mlFy5cwNtvv42UlBSrxFSxtWzZEn/++Sd+/vlnm/sNBgO++eYbPPTQQ2jXrp15e1RUFMaOHYsff/wRWq22zvPf+D1UVFTg8uXLuPPOOyEIAo4cOeJwfRtTn6efftqim+svf/kLDAYDzp07Z9dnbtu2Dbt27TK/TN2xBw8eRGFhIaZOnWqRFzR8+HB07twZX331ldW5nnvuOYv3H3/8MTQaDYYOHWrxf6V3795QqVTYu3cvAJhbgL788ktUV1fbVW8iezG4oWbJYDBgy5YtGDx4MHJzc3H69GmcPn0a/fr1Q0FBAXbv3m11TGpqKlq1amXx+vXXXy3KnDt3Drfccgt8fCz/a3Xp0sW8v7ZJkyZh3bp1AGq6BO68807ccsstVuWuXr2KefPmISYmBgqFAmFhYWjVqhWuXLmC0tJSq/JXrlxxKPDYtGkT/vvf/+K1115rsGxqaiqio6PxzDPP2HXu4uJi5OfnN+ql1+vtvgagpltNpVKhb9++uOWWWzBt2jTs37/fvP/SpUuorKxEp06drI7t0qULjEYjzp8/X+f5//jjD0ycOBEhISFQqVRo1aoVBg4cCAA2v4eGNKY+N998s8V7U5dS7ZyWutx1111ISEgwv/r37w/g+s+nrbp07tzZ6ue3RYsWuOmmmyy2/fbbbygtLUV4eLjV/5fy8nIUFhYCAAYOHIhHHnkECxYsQFhYGB588EGsW7cOOp3Ormsgqg9zbqhZ2rNnD/Ly8rBlyxZs2bLFav+mTZtwzz33WGx7+umn8dhjj1lsmzJlSpPrMm7cOMyaNQsHDhzAhg0bMHfuXJvlXnjhBaxbtw4vvvgi4uPjodFoIJPJMGbMGBiNRqvy+fn5SExMtKsOer0eKSkpeOqpp9CxY8d6y544cQLr16/Hxo0bIZfL7Tr/yJEjrVqt7LV3716HEpa7dOmCkydP4ssvv8TOnTuxbds2pKenY968eeZhxY1lMBgwdOhQFBcX4+WXX0bnzp0RGBiICxcuYOLEiTa/B1fw9fW1uV0QBLd8volCobAK5I1GI8LDw20m5wMwJwnLZDJ88sknOHDgAL744gt8/fXXePLJJ7FkyRIcOHDA41oEybswuKFmadOmTQgPDzePaLlRZmYmtm/fjjVr1lh0Qdxyyy1ISEiwKFt72GubNm3wn//8B0aj0eKXvqmFp02bNlafFxoaigceeADPPPMMCgsLMWrUKFy+fNmq3CeffIIJEyZgyZIl5m1VVVW4cuWKVdk///wTZWVl5hajhqSnp6OwsNCukShz5sxBjx49MHr0aLvODQBLliyxu1Whtri4OIePCQwMxOjRozF69Gjo9XqMHDkSr776KubMmYNWrVohICAAJ0+etDru119/hY+PD2JiYmye95dffsGpU6ewYcMGjB8/3rx9165dVmVrjzarS1Pq42ymn8+TJ09iyJAhFvtOnjxp8+e3tvbt2+Pbb79F//79Lf7/1OWOO+7AHXfcgVdffRWbN2/G448/ji1btmDy5Ml230Oi2hjcULNz9epVZGZm4rHHHsOjjz5qtT86OhoZGRn4/PPPHXqAA8CwYcPwzTffYOvWrea8m2vXruHtt9+GSqUyd1/U9uSTT2LYsGHm4bW2ghtfX1+rv8zffvttm/lBptao2g8oW8rKyvDqq68iOTkZkZGR9ZbNysrCN998g127djn04Ondu7fdZZuqqKgIoaGh5vd+fn7o2rUr/vWvf6G6uhpKpRL33HMPPvvsM5w9exaxsbEAgIKCAmzevBkDBgyAWq22eW5Ti8mN34MgCHjrrbesypoC3ytXrliMMLJ1zsbWx9n69OmD8PBwrFmzBk8++aQ5Z+tf//oXTpw4gXnz5jV4jlGjRiE9PR1paWlWXZzXrl1DeXk5WrZsiZKSErRs2dLi56hHjx4AYO6aMo2+shXAE9WHwQ01O59//jnKysrwwAMP2Nx/xx13oFWrVti0aZPDwc3TTz+Nd955BxMnTsShQ4cQGxuLTz75BPv378fy5csRFBRk87h7770Xly5dqrcp/v7778eHH34IjUaDrl27IisrC99++63Fg7ygoACpqan45z//iTFjxqBz584N1vnw4cMICwvDrFmzGiz7zTffYOjQoVYtWJ7knnvuQWRkJPr374+IiAicOHECK1euxPDhw833f+HChdi1axcGDBiAqVOnokWLFnjnnXeg0+nwj3/8o85zd+7cGe3bt8f//d//4cKFC1Cr1di2bZvNVilTQDd9+nQkJibC19cXY8aMsXnextbH2eRyOd544w1MmjQJAwcORFJSknkoeGxsrF1TCQwcOBDPPPMMFi1ahJycHNxzzz2Qy+X47bff8PHHH+Ott97Co48+ig0bNiA9PR0PP/ww2rdvj7KyMqxduxZqtRrDhg0DUJO83bVrV2zduhUdO3ZESEgIunfvju7du7v6VpC3E3OoFpEYRowYISiVSqGioqLOMhMnThTkcrlw+fJlh4aCC4IgFBQUCJMmTRLCwsIEPz8/4dZbb7UYfi0I9Q8vr2t/SUmJ+bwqlUpITEwUfv31V6FNmzbChAkTBEEQhP379wsdOnQQ5s+fL+h0OpvnrD0UHICwbNkyi7KmIcM3AiDIZDLh0KFDFttrDyd3N9QaCv7OO+8Id911lxAaGiooFAqhffv2wksvvSSUlpZaHHf48GEhMTFRUKlUQkBAgDB48GDhp59+sihjayj48ePHhYSEBEGlUglhYWHClClThKNHj1rd22vXrgkvvPCC0KpVK0Emk1ncT9gY3mxPfUxDwWsPn7ZVT1tM3+ulS5fqLbd161ahZ8+egkKhEEJCQoTHH39c+PPPPy3KTJgwQQgMDKzzHO+++67Qu3dvwd/fXwgKChJuvfVWYdasWcLFixfN15uUlCTcfPPNgkKhEMLDw4X7779fOHjwoMV5fvrpJ6F3796Cn58fh4WT3WSC4OYMNCIiIiIX4lBwIiIikhQGN0RERCQpDG6IiIhIUhjcEBERkaQwuCEiIiJJYXBDREREktLsJvEzGo24ePEigoKCOLU3ERGRlxAEAWVlZYiOjrZa06y2ZhfcXLx40W3rtBAREZFznT9/3mo1+tqaXXBjmn79/PnzbluvhYiIiJpGq9UiJiamzmVsbtTsghtTV5RarWZwQ0RE5GXsSSlhQjERERFJCoMbIiIikhQGN0RERCQpDG6IiIhIUhjcEBERkaQwuCEiIiJJYXBDREREksLghoiIiCSFwQ0RERFJSrOboZiISFKqSgFdOQxB0cjOLUZhWRXCg5To2zYEvmUXAYUKUGrEriWRWzG4ISLyVlWlwMZHUFmSj6TqFBzVqsy74tTlyJCnISA4Ehi3jQEONSvsliIi8la6clSW5COg4jxWVM1FFIoAAFEowoqquQioOI/KknxAVy5yRYnci8ENEZGXMgRFI6k6BeeM4WjjU4gtfmnoJTuFLX5paONTiHPGcCRVp8AQFC12VYncisENEZGXys4txlGtCmP01wOcTMV8c2AzRl/TVZWdWyx2VYncisENEZGXKiyrAgDkIRTJ1VMt9iVXT0UeQi3KETUXDG6IiLxUeJASQE2OzTJ5usW+ZfJ0cw6OqRxRc8HghojIS/VtG4I4dblFjs1I3XyLHJw4dTn6tg0Ru6pEbsXghojIS/mWXUSGPM0ix+aw0NEiBydDnlYz3w1RM8LghojIWylUCAiORGVgDKYrF5pzbPIQiunKhagMjKmZ50ahauBERNLCSfyIiLyVUgOM24YAXTkybc5QPIAzFFOzJGrLzffff48RI0YgOjoaMpkMn376ab3lMzMzMXToULRq1QpqtRrx8fH4+uuv3VNZIiJPpNQAmtbw9ZEhvn0oHuzRGvHtQ+HrIwM0rRnYULMkanBTUVGBuLg4rFq1yq7y33//PYYOHYodO3bg0KFDGDx4MEaMGIEjR464uKZERETkLWSCIAhiVwIAZDIZtm/fjoceesih47p164bRo0dj3rx5dpXXarXQaDQoLS2FWq1uRE2JiIjI3Rx5fnt1QrHRaERZWRlCQjjMkYiIiGp4dULx4sWLUV5ejlGjRtVZRqfTQafTmd9rtVp3VI2IqG5VpYCuHAabScAXmQRM1EReG9xs3rwZCxYswGeffYbw8PA6yy1atAgLFixwY82IiOpRVQpsfASVJflIqq5Z+8kkTl2ODHlazfDtcdsY4BA1kld2S23ZsgWTJ0/GRx99hISEhHrLzpkzB6WlpebX+fPn3VRLIiIbdOWoLMlHQMV5rKiaa14iIQpFWFE1FwEV51FZkg/oykWuKJH38rrgJiMjA5MmTUJGRgaGDx/eYHmFQgG1Wm3xIiISiyEoGknVKRZLJPSSnbJYQiGpOgWGoGixq0rktUQNbsrLy5GTk4OcnBwAQG5uLnJycvDHH38AqGl1GT9+vLn85s2bMX78eCxZsgT9+vVDfn4+8vPzUVpaKkb1iYgclp1bjKNalcUSCZmK+RZLKBzVqpCdWyx2VYm8lqjBzcGDB9GzZ0/07NkTADBz5kz07NnTPKw7Ly/PHOgAwLvvvotr165h2rRpiIqKMr9mzJghSv2JiBxVWFYFoGaJhOTqqRb7kqunmpdQMJUjIseJmlA8aNAg1DfNzvr16y3e79u3z7UVIiJysfAgJYCaHJtl8nSLfcvk6RijT0EeQs3liMhxXpdzQ0Tkzfq2DUGcutwix2akbr5FDk6cuhx923L+LqLGYnBDRORGvmUXkSFPs8ixOSx0tMjByZCn1cx3Q0SNwuCGiMidFCoEBEeiMjAG05ULzTk2eQjFdOVCVAbG1Mxzo1A1cCIiqovXTuJHROSVlBpg3DYE6MqRaXOG4gGcoZioiRjcEBG5m1IDKDXwBRDfPtRyn6a1KFUikhJ2SxEREZGkMLghIiIiSWFwQ0RERJLC4IaIiIgkhcENERERSQqDGyIiIpIUBjdEREQkKQxuiIiISFI4iR8ReQSDUbCerddHJna1iMgLMbghItHtPJaHBV8cR15plXlblEaJ1BFdcW/3KBFrRkTeiN1SRCSqncfy8NzGwxaBDQDkl1bhuY2HsfNYnkg1IyJvxeCGiERjMApY8MVxCDb2mbYt+OI4DEZbJYiIbGNwQ0Siyc4ttmqxuZEAIK+0Ctm5xe6rFBF5PQY3RCSawrK6A5vGlCMiAhjcEJGIwoOUTi1HRARwtBQRiahv2xBEaZTIL62ymXcjAxCpqRkWTtY4fJ7INgY3RCQaXx8ZUkd0xXMbD0MGWAQ4pkd06oiufGDbwOHzRHVjtxQRiere7lFYPa4XIjWWXU+RGiVWj+vFB7UNHD5PVD+23BCR6O7tHoWhXSPZxWKHhobPy1AzfH5o10i77h+7tkiKGNwQkUfw9ZEhvn2o2NXweI4Mn2/ofrJri6SK3VJERF7EWcPn2bVFUsbghog8jsEoIOtMET7LuYCsM0WcofgGzhg+z5mhSerYLUVEHoVdJfVzxvB5Z3ZtAczbIc/D4IaIPIapq6T2Q9vUVcLRU84ZPu/MmaEZjJInYrcUEXkEdpXYr6nD5501MzTzdshTseWGiDyCs7tK7OWtXSpNGT7vjK4tZw9JJ3ImBjdE5BHEWETT27tUGjt83hldW2IFo0T2YLcUEXkEdy6iaTAKeOvbU3i2GXepNLVriyu6kydjyw0ReQR3LaK581ge5n9+HPla2w/d5tSl0pSurcYEo97aBUjeh8ENkYj4y/46dyyiWddorNpMXSrr9+diYv+2kv5OGtu15Wgw6u1dgORdZIIgNKuhB1qtFhqNBqWlpVCr1WJXh5qxxvyybw7BkKseggajgAFv7Kk3T8SW+j679vfRu00wDp0rsfp+pPq9mYJFwHYwaureqiuorF3Oa1WVArpyGIKirb/nsouAQgUoNWLX0us58vxmcEMkgsb8sm9Of/nWFww0NlDIOlOEpLUHHK5LXd+Jre/DRwbcOFI9SqPEA3FR+PxonmS/t4Z+LhsKKk0tPD++PMQ7A76qUmDjI6gsyUdSdQqOalXmXXHqcmTI0xAQHAmM28YAp4kY3NSDwQ2JrTG/7CX/l6+dGhvgGYwCln5zEqv2nWnU58oAhAT6Ye7wLojU+KOkQo9pmxvu3qrvfIDj35untgDVVy97g8qMKXd456iq0guofDcRARXncc4YjjH6FOQhFFEowha/NLTxKURlYAwCnv4a0LQWu7ZezZHnN3NuiNzM0SG0Up9PxN4Htj2zFw/tGokDvxch60wRAAHx7cJQelWPv316DFcqqxtdRwFAUYUeyR8dBQCrnKDGnM/R782TW+7qy9uR+qgqQ1A0kqpTsMI4F218CrHFLw3J1VOxTJ6ONj6FOGcMx/TqFGQGRcNX7Mo2IwxuiNzM0V/2Up5PxN4Htj0B3uzMXzA78xeLIGbl3sa11DTEGc3djnxv3rwshTuH+IshO7cYR7UqjEGKuaUmUzEfAK635OhVXvn/05txnhsiN3P0l71U//J1ZOp+ewK8K5XVTWqdEUtD35u3L0thGlVVV9uUDDUBbVOH+IvF9P3lIRTJ1VMt9iVXT0UeQi3KkXswuCFyM0d/2UvxL9+GHtgCgFe2H4P+mhGAtB8MDX1vjrTceSLTEH8AVj/zzhriLybT9xeFIiyTp1vsWyZPRxSKLMqRezC4IXIzR3/ZS/Ev34Ye2EBNjssdi77FzmN5knww2PreDEYBWWeK8FnOBWSdKYLBKEii5a6psyF7sr5tQxCnLjd3SZ0zhmOkbj7OGcPNOThx6nKv+v8pBcy5IRKB6Zd97XyTSBv5Ju6Y3M7d7H0QF1dU47mNh7FqbM96J4zzNra+t7ryj8bcHmPXOc9ernB2NZ2qKbMhezLfsos1w731hRajpcbor+fgZMjT4Fs2gKOl3IhDwYlE5MjQXk8eLeMoR+ecidIokTK8K6Zttp4wzhvYmv/mxu+tvqH+AoCWAXKUVlY3eN1rvLwVxCtxnhu34Tw39WBwQ57I3iDHU+c5cZRprh9HWmKSE25Bp8ggqwAvJECOqmtGVOoNrqlsE8kArBrbE8GBijonJWxo3iONHcGN10+G5804Q7FbMLipB4Mb8jRSapFxhL3rPN1ozf/mssnOLca3x/OxPecCiitcP0IqJFDeqM+x53u0txXr0V6t8cnhCw2W89rJ8Iga4MjzmwnFRCJyZDi01JjyjkIC/ew+ZsEXxwEApVf1eH//WZcHNqak3wNzEpAyvItDxyYn3IIfXx7SYIBqb/5RgMK+FElPTiwmchdRg5vvv/8eI0aMQHR0NGQyGT799NMGj9m3bx969eoFhUKBDh06YP369S6vJ5ErePv8Jc5wb/coHJhzN0IC5XaVzyutwoHfi+q8byaBCl9MG9wem57qh/SxvRBVa5SOI502qSO6wq+FDyb2b2t3IJaccAtmJHS0q3vI3pFgbUIC7ConxZFlRI4SNbipqKhAXFwcVq1aZVf53NxcDB8+HIMHD0ZOTg5efPFFTJ48GV9//bWLa0rkfN4+f4mz+LXwwWsP32p3+awzRQ0OI6/QGTCgQyv0vyUMw26Lwo8vD0HGlDvw1pgeSBnexe6usMRuEbhQchXbDv2J9ftzMeK2yAaPidIo8fyQW+z8BPuH+j8RHyu5KQGIXEXUoeD33Xcf7rvvPrvLr1mzBm3btsWSJUsAAF26dMGPP/6IZcuWITEx0VXVJHIJKcxf4iz3do9CckJHLPv2lB2l7QtN6rpvJZV6u+u1878F2PnfArvLA8ADcVH1ttjYSgq3Z6i/XwsfyU0JQOQqXjXPTVZWFhISEiy2JSYm4sUXX6zzGJ1OB51OZ36v1WpdVT0ih3jjzMOuHK31/JAOyMg+h3ytzuZ+02ig+HZhdq0ZZbpvthK2Xenzo3mYdW+XOhf/rCt53J55jxyZH4moOfOq4CY/Px8REREW2yIiIqDVanH16lX4+/tbHbNo0SIsWLDAXVUkspupO6Ku4dCmh7mndDM4a1RXXQGSr48M8x/ohuc2Ws9lc2PLxB3tQ+26b73bBOOtb09h2be/OXytTVHXYpj2LH7548tDGgwepToZHpEzeVVw0xhz5szBzJkzze+1Wi1iYuyb8ZPIlbxp5mFnrUrdUIBUV8tEcKAcCx/sbv6Mhu7b/bdF4Y5Fu1FcYX8XlDO9891pADAHHfasar7gi+MY2jXSrmHcvj4yDvcmqodXDQWPjIxEQYFl/3dBQQHUarXNVhsAUCgUUKvVFi8iT+ENa+44a1SXvcPe7+0ehZThXS1GJhVXVCPtqxMWZeq6bwldw7H2h1zRAhsA2HfqMpLWHsCAN/Zg57E8Jo8TuZlXtdzEx8djx44dFtt27dqF+Ph4kWpE1HSe3s3gyIO5rtYER1oudh3Px7TN1q1EebVaiWzdt0ulVZj+UU7jLtQFTIHbpP6xdpVvDsnjTSWVWbrJtUQNbsrLy3H69Gnz+9zcXOTk5CAkJAQ333wz5syZgwsXLuCDDz4AADz77LNYuXIlZs2ahSeffBJ79uzBRx99hK+++kqsSyByCk/uZnDGqC57A6Sffrtc7xw2AoA5mb9gaNdIc56O6b7t+E8eZnycY1dd3cUUuH2Wc9Gu8p6UPO6Jmuts3uQ4UbulDh48iJ49e6Jnz54AgJkzZ6Jnz56YN28eACAvLw9//PGHuXzbtm3x1VdfYdeuXYiLi8OSJUvwz3/+k8PAiVzIGaO67A2Qnt10qMFRTSWV1Vi5xzJJeOexPEzdfBieuJiMAKCoQo+QQDnnqGmC5jybNzlO1JabQYMGob6lrWzNPjxo0CAcOXLEhbUiohs5Y1SXvQFShZ2LX67bfxbPD7nFIlnX0z3cozXe33/WruRxdr1YcqRbsznfJ7rOq3JuiMj9TKO6nv3fEO3aBDQ8qquhAMlRV65W48CZIvj4yLD/9CW3zWHTFAldI3F725AG56hh14s1Z+R9UfPC4IaIXO7GYe/O8vSHB+1u6RFbywC5ufXFlASdr61CcbkOIYF+0Pj7wWAU8PWxPEzdbN0y7eiQe6nhbN7kKAY3RFSvhrp97O0SMA3fnr3tF1y52vTVvD0psFG08IHumrHO/TfeFV8fGUqv6vGPnb9atEa09G+B0qprNo9v7l0v3jibN4nLq+a5ISL3c+YcLfd2j8Kqx3s5sXbiCw30qzewAWqSoE33p67E2CtXr9WbEN2c58Kxd3FRJmSTCYMbIqqXs7sE7mgX2uCDKjhAbl/lRCT73+vBHtF2lS8sq6o3MdZezbHrxdStCcDq58bTZvNutqpKgdILMBgFZJ0pwmc5F5B1pqhmcs/SCzX73YjdUkRUL2d3Cdiz7MTEO9ui9Koen+ZcFHWmYZOW/wu2rlRe704zJQJr/P3w/v6zDZ4jPEjZYCuYPZpr1wsXDfVgVaXAxkdQWZKPpOoUHNWqzLvi1OXIkKchIDgSGLcNUGrcUiUGN0RUL1cs8FnXg6plgBwCgGXfnjJvC1T4okLn/vyakEA5Uu7vhkj19WuzNTzbYBTsvj9f/se+yfzq0ty7Xjx9Nu9mS1eOypJ8BFScxwrjXIxBCvIQiigUYUVVGgL0hagEEKArZ3BDRJ7BVQt81n5Qnb1cYXMFbzECGxmA1x6+1ao1oPYwY9N8NMO6R+I9G603te9PU1td2PXi2bN5N1eGoGgkVadghXEu2vgUYotfGpKrp2KZPB1tfApxzhiO6dUpyAyKhq+b6sTghshD2TORm7sme3N2l0Dtet/XPQoD39zr7Go3SssAOV4faR3Y1GZrPhofGXDj+qG1709j5/vxkQErk3qy64U8UnZuMY5qVRiDFGzxS0Mbn0JkKuYDAM4ZwzFGn4I8vcqt8xAxuCHyQPZM5Obuyd6c1SVgq94hgX4ekVsDAKWVDQ9TN414qh2gmEY7Pdk/FkO7Rlrdn/paweqzMqkXht3GwIY8kynJPQ+hSK6eag5sACC5eiryEGpRzh04WorIw9izho5Y6+yYugQe7NEa8e1DGxXY2Kq3vYFNgJ97GrUXfHG8ZpSHDfYsBfCvY/l1Bn6mVrBIjWUXVcsAuTlx2SRKo8SacQxsyLOZulujUIRl8nSLfcvk6YhCkUU5d2DLDZEHsXcNHUEQvG6dHWcMg177RB8cPFeMdfvPOmUiQFsamsrfGUsB1NUKZjo/k2XJm/RtG4I4dTlWVKWZc2xuzLnZ4peG6cqFbk2GZ3BD5EHsfXDWx1PX2WnKMGjTiKM72oei/y1heH7ILTckIlciI/sP5Gud2+RdVxO6s+b9qSsx1pO+MyJ7+JZdrBnurS+8nmODUIzRX8/ByZCnwbdsAKBp7ZY6MbghagRXJfI6s0/aVf3bjb32xtbH1ois2oHB80M6YNmuU1i593SjPsOWy2U6fJZzAWGBCkAGXC7XITxIiTCVwq7jm+t8NNQMKVQICI5EJYDp1TXJw0BNDs505cLr89woVPWfx4kY3BA5yJWJvM58ILri4dqUa7e3PiGBchRXWE+WV9/5fX1k6N8hzKnBTdpXJ2xuj1Qr0DJAjtLKaqfN+0Pk1ZQaYNw2BOjKkRkUbf3HT9mAmsDGTXPcAAxuiBxS1ygZZ63abO+EeYIgoECrc+vDtanXbu+1fffSYBw6V+Jwy5A9w6xrD9VujBvve1Pn/XHXUH4il1NqAKUGvrDRteqmrqgbcbQUkZ0aSvYF6h9lYw9719CZ/0C3Bss48yHpjGu399r8Wvg0akRWQ+eXoWZI9bRB7e06X11MSdstA+SIUFu2RkVqlHYHuDuP5WHAG3uQtPYAZmzJQdLaAxjwxh6XjXQjak4Y3BDZyZmrY9enrqHCNz447SnjTM66dlfXu6HzD7stCiGBfk36DKDmeq9UVmPJY3HImHIH3hrTAxlT7sCPLw+xO7ARYyg/UXPBbikiOzl7dez62DNhnjvX2XHmtbu63g2dv6W/81Ycv1yhw4M9HGtyt3e4v6cN5SfyJgxuiOzk7NWxG2LPGjruWmfHFSuDu7Le9Z3fmfPjmK7XkdwZZ8yTQ0T1Y3BDZCdXrI7tLRy9dk9OlA2xcyh3fW68XkdHkNnbCvbt8XwGN0SNxJwbIjvZmxDryEPcYBSQdaYIn+VcQNaZoiYlI7uSI9fu6YmykWrntKyljuiKXcfzHc6dsbd16739Zz3mnhF5GwY3RA5wZkKspwcBtdlz7d6QKNu3bUiTkoqj/ne9Q7tGNmoEmakVzB5NHX1H1FzJBEFoVv9ztFotNBoNSktLoVarxa4OeammdrvUNWeM6QyuGPHkLHVdu8EoYMAbe+rMJzF15fz48hDRu6h2/CcPUzcfdvi45IRb8PyQW+DrI0PWmSIkrT3Q4DEZU+5A37YhFvfscrkOL2QcseszM6bcwe4pIjj2/GbODVEjNCUh1ttHy9R17d6UKDvstig882dbvPN9rl3lbeXQOJI7M/OjHIt7ExJo/4gtVy2jQSRlDG6I3MyTggBnJv66c6i8M8wZ1hVxN7XE3M+OWSz3EKVRImV4VwQH+tV7XxzJnantxs9rCNeoInIcgxsiN/OUIMDZa2S5e6i8Mwy7LRqJ3aMaFeDZM4JM1sTlHkIC5ZIcfUfkakwoJnIzTwgCXJH4a3rY1xUWyFATPHnaw9rUzebs5R4ENH0dq4UPdvfIrkkiT8fghsjNxA4CXLVGliuGynu6+kaQPdk/tknnfuauthh2W3STzkHUXLFbisjNTEHAcxsPN3lV6cZwZc6P6WFfu7srsgndXZ6uruUeDpwpwvs28m1qCwn0Q3GF3vw+NNAPaQ92x7DbpHeviNyFwQ2RCFwRBNibHOzqnB93rnnVVM5KqK49gmznsTzM//x4vceYhsZ/99JgHDpX4vH3isibMLhpqqpSQFcOQ1C09S/JsouAQgUoNWLXkjyQM4MAR5KD3ZHz4641r5rC1j0LCfTDQz2iMbRrZJO+C1tzGN3oxhY6vxY+Hn+viLwNJ/FriqpSYOMjqCzJR1J1Co5qVeZdcepyZMjTEBAcCYzbxgCHXMbRCQFNk+01tE6UJ0y25yr2BCCNGTnW0ESGJpFqBeY/0E2S3XREruLI85sJxU2hK0dlST4CKs5jRdVcRKEIABCFIqyomouAivOoLMkHdOUiV5SkqjHJwc0x8fdG9d2zG+U1YuRYQ/lMJktG9WBgQ+RCDG6awBAUjaTqFJwzhqONTyG2+KWhl+wUtviloY1PIc4Zw5FUnQJDEEc8kGs4khx8I2eukeWJ6luQ1N4AxMSRkWP25illnbnskWtGectCrkQNYc5NE2TnFuOoVoUxSDEHNJmK+QCAc8ZwjNGnIE+v8ojp5kmampIcXFfODwBknSny2gTXhvKPHEmUdnTkmL15Siv3nsG2wxc8agSZsyd1JBITW26awPRLMg+hSK6earEvuXoq8hBqUY7I2ZqSHGxrpNCu4/letVJ5bfZMTtiYRGl7/w83NIdRXXUSmzes5k7kCAY3TWD6JRmFIiyTp1vsWyZPN+fgeNJ08yQtjZ0QcOexPKsgpvfCXXjWix9w9uYf9W4TjCiNY/8n7f0/XF8+U311ErP7x1WTOhKJicFNE/RtG4I4dblFjs1I3XyLHJw4dbnHTTdP0tGY5OC6/kq/Uml7MUdvecDZm3906FwJHoizr5ulMbNF15XPVF+daudEuVNj87aIPBmDmybwLbuIDPn1wGaMPgWHhY4Yo7+eZJwhT6uZ74bIRRxJDrZ3pFBt3vCAs7frKF9bhc+PNtwK1ZSRY/d2j8KPLw/B84M72FVezK5rT1nIlciZmFDcFAoVAoIjUQlgenVN8jBQk4MzXbnw+jw3ClX95yFqInsnBHR0pFBtnvyAs7frqLhcZ9c9CAn0w6sPd290Mq2vjwz9O4Rh5d7TDZYVs+vaExZyJXI2BjdNodQA47YhQFeOTJszFA/gDMXkNvbMCtzU4MSTH3Cm/KOGJicMCfSz63xzh3dp8ighe+skZte1N9SRyFHslmoqpQbQtDY/WB7s0Rrx7UNr/mLWtGZgQx6lscGJq1cqdwZ7848iNf52nc/ecs6ok5hD7b2hjkSOYnBD1Iw4MlTZROwHnCMTy9mTf9TYEWaN5Q0TJnpDHYkcwbWliJoZ02gpABbdELL/vW8ZILcYOSXmRG6NnViuodW+67sHgPV6XM7grBXIXckb6kjNlyPPbwY3RM1QfUGDs1Yqd0YdHVkQtDHn54y8RN6DwU09GNwQ1fDkv9IbWl3bWSuXe/I9ICJLXrUq+KpVqxAbGwulUol+/fohOzu73vLLly9Hp06d4O/vj5iYGCQnJ6OqynOHpxJ5KptJ8B7CXRPLefI9IKLGEzW42bp1K2bOnInU1FQcPnwYcXFxSExMRGFhoc3ymzdvxuzZs5GamooTJ07gvffew9atW/G3v/3NzTUnIlfixHJE1BSiBjdLly7FlClTMGnSJHTt2hVr1qxBQEAA3n//fZvlf/rpJ/Tv3x9jx45FbGws7rnnHiQlJTXY2iNJVaVA6QXbI0lKL9TsJ/JSnFiOiJpCtEn89Ho9Dh06hDlz5pi3+fj4ICEhAVlZWTaPufPOO7Fx40ZkZ2ejb9+++P3337Fjxw488cQTdX6OTqeDTqczv9dqtc67CLFUlQIbH0FlST6SqlNwVHt9BuQ4dfn1mZHHbeM8O+SVOLEcETWFaC03ly9fhsFgQEREhMX2iIgI5Ofn2zxm7Nix+Pvf/44BAwZALpejffv2GDRoUL3dUosWLYJGozG/YmJinHodotCVo7IkHwEV57Giaq559fEoFGFF1VwEVJxHZUk+oCsXuaJEjeOuieUcmUOHiLyH6AnFjti3bx9ee+01pKen4/Dhw8jMzMRXX32FtLS0Oo+ZM2cOSktLza/z58+7scauYQiKRlJ1isXq471kpyxWJ0+qToEhKFrsqhI1mqsnltt5LA8D3tiDpLUHMGNLDpLWHsCAN/Zg57GGF9UkIs8mWrdUWFgYfH19UVBQYLG9oKAAkZGRNo9JSUnBE088gcmTJwMAbr31VlRUVODpp5/GK6+8Ah8f61hNoVBAoVA4/wJElJ1bjKNaFcYgxRzQZCrmA4B5dfI8vQrZucUNrjVE5MnsXRDUUXXNoZNfWoXnNh7mrLxEXk60lhs/Pz/07t0bu3fvNm8zGo3YvXs34uPjbR5TWVlpFcD4+voCAJrTdD2mESJ5CEVy9VSLfcnVU5GHUItyTsdkZnIjZw/XNhgFLPjiuM1cHtO2BV8cZxcVkRcTdVXwmTNnYsKECejTpw/69u2L5cuXo6KiApMmTQIAjB8/Hq1bt8aiRYsAACNGjMDSpUvRs2dP9OvXD6dPn0ZKSgpGjBhhDnKaA9MIkSgUYZk83WLfMnl6TcsNQl0zkoTJzNRInjJhniNz6LDlk8g7iRrcjB49GpcuXcK8efOQn5+PHj16YOfOneYk4z/++MOipWbu3LmQyWSYO3cuLly4gFatWmHEiBF49dVXxboEUfRtG4I4dTlWVF3PsUmunopl8nRzDs505ULXjCS5MZnZOBdjUBNI1SQzpyFAX4hKAAG6cgY3ZOZJSx1wDh0i6ePyC96o9AIq301EQMX56zk2/wswTDk4lYExCHj6a0DT2qkfbTAKGPn6R1hRNddmYHXOGI7pyoXInD2Ks70SANevEeWorDNFSFp7oMFyGVPuYMsNkQfxquUXqBEUKgQER6IyMAbTlQvNOTZ5CMV05cKawCY4ElCoGjiR48zJzPrro7UyFfPNgc0YfU1XVVOnxSdp8MT8FtMcOnWF3jLUtCpxDh0i7yVqtxQ1klIDjNuGAF05MoOirfMYygbUBDYu6BaqncxsGqUFuCmZmbyKJ+a3mObQeW7jYcgAi8DLmXPoEJF42HLjrZQaQNPa9kgSTWuX5bs0lMxsmlDQ66bF5wgwl/DU/BZXz6HjCpxwkMh+bLkhh4iazOwqHAHmMp68RpSr5tBxhtojy0oqdEj76oRHJGQTeQMGN+QQ37KLNQ97faFFMvMY/fUJBTPkaTVdY05OZnYZjgBzGU9fI8rU8ulJbI0ss4UTDhLVjd1S5BgRk5ldhctZuI671oiSCtPIsoYCG4ATDhLVh0PByXFVpYCuHAabycwXXZbM7CqmocE3DqU3ubF1ikODG8+T5rnxVAajgAFv7LErsKmNP5vUHDjy/Ga3FDlOqQGUGvgC1r9QvaUr6gYcAeZ6YuW3eMqsyPZoaGRZffizSWSJwQ01e6IuZ9GMuDu/xdtai5oSoPBnk8gSc26o2TONALsxx2akbr5FDk6cuty7RoA1c3XlrpiScHceyxOpZnVrTIDCCQeJbGNwQ82eaQTYjbMsHxY6WszCXDMC7KLYVSU7eOKsyPZoaObk2piQTVQ3BjdEEhwB1pw5MiuyJ6lvZJktnjzhIJHYmHNDJOJyFuR8njorsj1MMyfbyhVKGd4VwYF+XpEcTSQ2BjdEgORGgDVnnjwrsj08eeZkIm/B4IaIJMXTZ0W2hyfOnEzkTZhzQ0SSwlmRiYjBDRFJjjeu+k1EzsNuKSKSJOauEDVfDG6ISLKYu0LUPLFbioiIiCSFwQ0RERFJCoMbIiIikhQGN0RERCQpDgc3Wq0WRqPRarvBYIBWq3VKpYiIiIgay6HgZvv27ejTpw+qqqzXZKmqqsLtt9+OL774wmmVIyIiInKUQ8HN6tWrMWvWLAQEBFjtCwwMxMsvv4yVK1c6rXJERNQIVaVA6QUYjAKyzhThs5wLyDpTBINRAEov1OwnkjCH5rk5duwY0tPT69x/1113Ye7cuU2uFBERNVJVKbDxEVSW5COpOgVHtSrzrjh1OTLkaQgIjgTGbeNK9yRZDrXclJSU4Nq1a3Xur66uRklJSZMrRUREjaQrR2VJPgIqzmNF1VxEoQgAEIUirKiai4CK86gsyQd05SJXlMh1HApuYmNjcfDgwTr3Hzx4EG3atGlypYiIqHEMQdFIqk7BOWM42vgUYotfGnrJTmGLXxra+BTinDEcSdUpMARFi11VIpdxKLgZOXIkXnnlFRQUFFjty8/Px9y5c/HII484rXJEROSY7NxiHNWqMEZ/PcDJVMw3BzZj9DVdVdm5xWJXlchlHMq5mT17Nj777DPccsstGDduHDp16gQA+PXXX7Fp0ybExMRg9uzZLqkoERE1rLCsZjRrHkKRXD0VmYr55n3J1VORh1CLckRS5FBwExQUhP3792POnDnYunWrOb+mZcuWGDduHF599VUEBQW5pKJERNSw8CAlgJocm2VyywEgy+TpGKNPQR5CzeWIpMjhSfw0Gg3S09Nx+fJlFBQUID8/H0VFRUhPT0dwcLAr6khERHbq2zYEcepyixybkbr5Fjk4cepy9G0bInZViVzGoZYbADhw4AC++OILVFdXY8iQIbj33ntdUS8iImoE37KLNcO99ddzbPIQijH6FHPAkyFPg2/ZAEDTWuzqErmEQy03n3zyCfr374+33noLa9euxfDhw7F48WJX1Y2IiBylUCEgOBKVgTGYrlxozrHJQyimKxeiMjCmZp4bhaqBExF5L5kgCIK9hXv37o3bb78dq1atgq+vLxYtWoQ333wTxcXek3Wv1Wqh0WhQWloKtVotdnWIiJyvqhTQlcMQFI3s3GIUllUhPEiJvm1D4Ft2sSaw4QR+5GUceX47FNyoVCrk5OSgQ4cOAAC9Xo/AwEBcuHAB4eHhTau1mzC4ISIi8j6OPL8d6paqrKy0OKGfnx+USiXKyznTJREREXkGhxOK//nPf0Klut5Xe+3aNaxfvx5hYWHmbdOnT3dO7YiIiIgc5FC3VGxsLGQyWf0nlMnw+++/N7lirsJuKSIiIu/jyPPboZab//znPw2e8Oeff3bklERERERO5VDOTWJiYp35NdeuXUNKSgr69+/vlIoRERERNYZDwU1ZWRkSEhKg1Wotth87dgy333473n//fXz66afOrB8RERGRQxwKbvbu3YuKigoMHToUWq0WgiDgjTfeQJ8+fdClSxccO3YMw4YNc1VdiYiIiBrkUM5Nq1atsGfPHiQkJGDIkCFQKBT47bffsHHjRjz66KOuqiMReQJODEdEXsLhoeCtWrXC7t27kZCQgGPHjiEnJwedO3d2Rd2IyFNUlQIbH0FlST6SqlNwVHt9Oog4dXnNWkbBkcC4bQxwiEh0Dq8KDgBhYWHYs2cPunbtirFjx6KkpMTZ9SIiT6IrR2VJPgIqzmNF1VxEoQgAEIUirKiai4CK86gsyQd0nNCTiMTnUHAzcuRI82vy5MlQq9XIyclB3759LfY5YtWqVYiNjYVSqUS/fv2QnZ1db/krV65g2rRpiIqKgkKhQMeOHbFjxw6HPpOIHGMIikZSdQrOGcPRxqcQW/zS0Et2yrzK9DljOJKqU2AIiha7qkREjnVLaTQaq/dt27Zt9Idv3boVM2fOxJo1a9CvXz8sX74ciYmJOHnypM21qvR6PYYOHYrw8HB88sknaN26Nc6dO4eWLVs2ug5E1LDs3GIc1aowBinmgCZTMR8AcM4YjjH6FOTpVcjOLUZ8+1BxK0tEzZ5Dwc26deuc+uFLly7FlClTMGnSJADAmjVr8NVXX+H999/H7Nmzrcq///77KC4uxk8//QS5XA6gZtZkInKtwrIqAEAeQpFcPdUc2ABAcvVU5CHUohwRkZgalXPjDHq9HocOHUJCQsL1yvj4ICEhAVlZWTaP+fzzzxEfH49p06YhIiIC3bt3x2uvvQaDweCuahM1S+FBSgA1OTbL5OkW+5bJ0805OKZyRERiEi24uXz5MgwGAyIiIiy2R0REID8/3+Yxv//+Oz755BMYDAbs2LEDKSkpWLJkCRYuXFjn5+h0Omi1WosXETmmb9sQxKnLLXJsRurmW+TgxKnL0bdtiNhVJSISL7hpDKPRiPDwcLz77rvo3bs3Ro8ejVdeeQVr1qyp85hFixZBo9GYXzExMW6ssRerKgVKL8BgFJB1pgif5VxA1pkiGIwCUHqhZj81G75lF5Ehvx7YjNGn4LDQEWP015OMM+RpNfPdEBGJzOF5bpwlLCwMvr6+KCgosNheUFCAyMhIm8dERUVBLpfD19fXvK1Lly7Iz8+HXq+Hn5+f1TFz5szBzJkzze+1Wi0DnIZwThOqTaFCQHAkKgFMr65JHgZqcnCmKxde/5lQqOo/DxGRG4gW3Pj5+aF3797YvXs3HnroIQA1LTO7d+/G888/b/OY/v37Y/PmzTAajfDxqWl0OnXqFKKiomwGNgCgUCigUChccg2SdeOcJsa5GIMU5CH0f3OapCFAX4hKAAG6cgY3zYVSA4zbhgBdOTJtzlA8gDMUE5HHELVbaubMmVi7di02bNiAEydO4LnnnkNFRYV59NT48eMxZ84cc/nnnnsOxcXFmDFjBk6dOoWvvvoKr732GqZNmybWJUgS5zQhm5QaQNMavj4yxLcPxYM9WiO+fSh8fWSApjUDGyLyGKK13ADA6NGjcenSJcybNw/5+fno0aMHdu7caU4y/uOPP8wtNAAQExODr7/+GsnJybjtttvQunVrzJgxAy+//LJYlyBJnNOE3IbrVRGRC8gEQRDEroQ7abVaaDQalJaWQq1Wi10dj/RZzgXM2JIDAOglO2Uxp8lI3XwcFjoCAN4a0wMP9mgtQg1JEpjbRUQOcOT57VWjpcg9OKcJuQXXqyIiF2FwQ1Y4pwm5A3O7iMhVGNyQFc5pQu5gzu264ecqUzHf4ufuqLYmt4uIyBEMbsiaaU6TwBhMVy40rxtkmtOkMjCGc5pQk9Ver+pGXK+KiJpC1NFS5KE4pwm5QUO5XWP0NfMrMbeLiBzFlhuyjXOakIsxt4uIXIXBDRGJgrldROQqDG6ISBzM7SIiF2HODRGJg7ldROQiDG6ISDxKDaDUwBewXspDw9mviahx2C1FREREksLghoiIiCSFwQ0RERFJCoMbIiIikhQGN0RERCQpDG6IiIhIUhjcEBERkaQwuCEiIiJJYXBDREREksLghoiIiCSFwQ0RERFJCoMbIiIikhQGN0RERCQpDG6IiIhIUhjcEBERkaQwuCEiIiJJYXBDREREksLghoiIiCSFwQ0RERFJCoMbIiIikhQGN0RERCQpDG6IiIhIUhjcEBERkaQwuCEiIiJJYXBDREREksLghoiIiCSFwQ0RERFJCoMbIiIikhQGN0RERCQpDG6IiIhIUhjcEBERkaQwuCEiIiJJYXBDREREksLghoiIiCSFwQ0RERFJCoMbIiIikhQGN0RERCQpDG6IiIhIUjwiuFm1ahViY2OhVCrRr18/ZGdn23Xcli1bIJPJ8NBDD7m2gkREROQ1RA9utm7dipkzZyI1NRWHDx9GXFwcEhMTUVhYWO9xZ8+exf/93//hL3/5i5tqSkRERN5A9OBm6dKlmDJlCiZNmoSuXbtizZo1CAgIwPvvv1/nMQaDAY8//jgWLFiAdu3aubG2RERE5OlEDW70ej0OHTqEhIQE8zYfHx8kJCQgKyurzuP+/ve/Izw8HE899VSDn6HT6aDVai1eREREJF2iBjeXL1+GwWBARESExfaIiAjk5+fbPObHH3/Ee++9h7Vr19r1GYsWLYJGozG/YmJimlxvIiIi8lyid0s5oqysDE888QTWrl2LsLAwu46ZM2cOSktLza/z58+7uJZEREQkphZifnhYWBh8fX1RUFBgsb2goACRkZFW5c+cOYOzZ89ixIgR5m1GoxEA0KJFC5w8eRLt27e3OEahUEChULig9kREROSJRG258fPzQ+/evbF7927zNqPRiN27dyM+Pt6qfOfOnfHLL78gJyfH/HrggQcwePBg5OTksMuJiIiIxG25AYCZM2diwoQJ6NOnD/r27Yvly5ejoqICkyZNAgCMHz8erVu3xqJFi6BUKtG9e3eL41u2bAkAVtuJiIioeRI9uBk9ejQuXbqEefPmIT8/Hz169MDOnTvNScZ//PEHfHy8KjWIiIiIRCQTBEEQuxLupNVqodFoUFpaCrVaLXZ1iIiIyA6OPL/ZJEJERESSwuCGiIiIJIXBDREREUkKgxsiIiKSFAY3REREJCkMboiIiEhSGNwQERGRpDC4ISIiIklhcENERESSwuCGiIiIJIXBDREREUkKgxsiIiKSFAY3REREJCkMboiIiEhSGNwQERGRpDC4ISIiIklhcENERESSwuCGiIiIJIXBDREREUkKgxsiIiKSFAY3REREJCkMboiIiEhSGNwQERGRpDC4ISIiIklhcENERESS0kLsChARuV1VKaArhyEoGtm5xSgsq0J4kBJ924bAt+wioFABSo3YtSSiRmJwQ0TNS1UpsPERVJbkI6k6BUe1KvOuOHU5MuRpCAiOBMZtY4BD5KXYLUVEzYuuHJUl+QioOI8VVXMRhSIAQBSKsKJqLgIqzqOyJB/QlYtcUSJqLAY3RNSsGIKikVSdgnPGcLTxKcQWvzT0kp3CFr80tPEpxDljOJKqU2AIiha7qkTUSAxuiKhZyc4txlGtCmP01wOcTMV8c2AzRl/TVZWdWyx2VYmokRjcEFGzUlhWBQDIQyiSq6da7Euunoo8hFqUIyLvw+CGiJqV8CAlgJocm2XydIt9y+Tp5hwcUzki8j4MboioWenbNgRx6nKLHJuRuvkWOThx6nL0bRsidlWJqJEY3BBRs+JbdhEZ8jSLHJvDQkeLHJwMeVrNfDdE5JUY3BBR86JQISA4EpWBMZiuXGjOsclDKKYrF6IyMKZmnhuFqoETEZGn4iR+RNS8KDXAuG0I0JUj0+YMxQM4QzGRl2NwQ0TNj1IDKDXwBRDfPtRyn6a1KFUiIudhtxQRERFJCoMbIiIikhQGN0RERCQpDG6IiIhIUhjcEBERkaQwuCEiIiJJYXBDREREksLghoiIiCSFwQ0RERFJCoMbIiIikhSPCG5WrVqF2NhYKJVK9OvXD9nZ2XWWXbt2Lf7yl78gODgYwcHBSEhIqLc8ERERNS+iBzdbt27FzJkzkZqaisOHDyMuLg6JiYkoLCy0WX7fvn1ISkrC3r17kZWVhZiYGNxzzz24cOGCm2tOREREnkgmCIIgZgX69euH22+/HStXrgQAGI1GxMTE4IUXXsDs2bMbPN5gMCA4OBgrV67E+PHjGyyv1Wqh0WhQWloKtVrd5PoTERGR6zny/Ba15Uav1+PQoUNISEgwb/Px8UFCQgKysrLsOkdlZSWqq6sREhJic79Op4NWq7V4ERERkXSJGtxcvnwZBoMBERERFtsjIiKQn59v1zlefvllREdHWwRIN1q0aBE0Go35FRMT0+R6ExERkecSPeemKV5//XVs2bIF27dvh1KptFlmzpw5KC0tNb/Onz/v5loSERGRO7UQ88PDwsLg6+uLgoICi+0FBQWIjIys99jFixfj9ddfx7fffovbbrutznIKhQIKhcIp9SUiIiLPJ2rLjZ+fH3r37o3du3ebtxmNRuzevRvx8fF1HvePf/wDaWlp2LlzJ/r06eOOqhIREZGXELXlBgBmzpyJCRMmoE+fPujbty+WL1+OiooKTJo0CQAwfvx4tG7dGosWLQIAvPHGG5g3bx42b96M2NhYc26OSqWCSqUS7TqIiIjIM4ge3IwePRqXLl3CvHnzkJ+fjx49emDnzp3mJOM//vgDPj7XG5hWr14NvV6PRx991OI8qampmD9/vjurTkRERB5I9Hlu3I3z3BAREXkfr5nnhoiIiMjZGNwQERGRpDC4ISIiIklhcENERESSwuCGiIiIJIXBDREREUkKgxsiIiKSFAY3REREJCkMboiIiEhSGNwQERGRpDC4ISIiIklhcENERESSwuCGiIiIJIXBDREREUkKgxsiIiKSFAY3REREJCkMboiIiEhSGNwQERGRpDC4ISIiIklhcENERESSwuCGiIiIJIXBDREREUlKC7ErQEQuUFUK6MphCIpGdm4xCsuqEB6kRN+2IfAtuwgoVIBSI3YtiYhcgsENkdRUlQIbH0FlST6SqlNwVKsy74pTlyNDnoaA4Ehg3DYGOEQkSeyWIpIaXTkqS/IRUHEeK6rmIgpFAIAoFGFF1VwEVJxHZUk+oCsXuaJERK7B4IZIYgxB0UiqTsE5Yzja+BRii18aeslOYYtfGtr4FOKcMRxJ1SkwBEWLXVUiIpdgcEMkMdm5xTiqVWGM/nqAk6mYbw5sxuhruqqyc4vFrioRkUswuCGSmMKyKgBAHkKRXD3VYl9y9VTkIdSiHBGR1DC4IZKY8CAlgJocm2XydIt9y+Tp5hwcUzkiIqlhcEMkMX3bhiBOXW6RYzNSN98iBydOXY6+bUPErioRkUswuCGSGN+yi8iQp1nk2BwWOlrk4GTI02rmuyEikiAGN0RSo1AhIDgSlYExmK5caM6xyUMopisXojIwpmaeG4WqgRMREXknTuJHJDVKDTBuGwJ05ci0OUPxAM5QTESSxuCGSIqUGkCpgS+A+Pahlvs0rUWpEhGRu7BbioiIiCSFwQ0RERFJCoMbIiIikhQGN0RERCQpDG6IiIhIUhjcEBERkaQwuCEiIiJJYXBDREREksLghoiIiCSl2c1QLAgCAECr1YpcEyIiIrKX6blteo7Xp9kFN2VlZQCAmJgYkWtCREREjiorK4NGU//aeDLBnhBIQoxGIy5evIigoCDIZDKxq+NVtFotYmJicP78eajVarGrI2m81+7De+0+vNfuJbX7LQgCysrKEB0dDR+f+rNqml3LjY+PD2666Saxq+HV1Gq1JP6jeAPea/fhvXYf3mv3ktL9bqjFxoQJxURERCQpDG6IiIhIUhjckN0UCgVSU1OhUCjErork8V67D++1+/Beu1dzvt/NLqGYiIiIpI0tN0RERCQpDG6IiIhIUhjcEBERkaQwuCEiIiJJYXBDFlatWoXY2FgolUr069cP2dnZ9Za/cuUKpk2bhqioKCgUCnTs2BE7duxwU229m6P3evny5ejUqRP8/f0RExOD5ORkVFVVuam23uv777/HiBEjEB0dDZlMhk8//bTBY/bt24devXpBoVCgQ4cOWL9+vcvrKQWO3uvMzEwMHToUrVq1glqtRnx8PL7++mv3VNbLNebn2mT//v1o0aIFevTo4bL6iY3BDZlt3boVM2fORGpqKg4fPoy4uDgkJiaisLDQZnm9Xo+hQ4fi7Nmz+OSTT3Dy5EmsXbsWrVu3dnPNvY+j93rz5s2YPXs2UlNTceLECbz33nvYunUr/va3v7m55t6noqICcXFxWLVqlV3lc3NzMXz4cAwePBg5OTl48cUXMXnyZD507eDovf7+++8xdOhQ7NixA4cOHcLgwYMxYsQIHDlyxMU19X6O3muTK1euYPz48bj77rtdVDMPIRD9T9++fYVp06aZ3xsMBiE6OlpYtGiRzfKrV68W2rVrJ+j1endVUTIcvdfTpk0ThgwZYrFt5syZQv/+/V1aT6kBIGzfvr3eMrNmzRK6detmsW306NFCYmKiC2smPfbca1u6du0qLFiwwPkVkjBH7vXo0aOFuXPnCqmpqUJcXJxL6yUmttwQgJpWmEOHDiEhIcG8zcfHBwkJCcjKyrJ5zOeff474+HhMmzYNERER6N69O1577TUYDAZ3VdsrNeZe33nnnTh06JC56+r333/Hjh07MGzYMLfUuTnJysqy+G4AIDExsc7vhpzHaDSirKwMISEhYldFktatW4fff/8dqampYlfF5Zrdwplk2+XLl2EwGBAREWGxPSIiAr/++qvNY37//Xfs2bMHjz/+OHbs2IHTp09j6tSpqK6ubhb/eRqrMfd67NixuHz5MgYMGABBEHDt2jU8++yz7JZygfz8fJvfjVarxdWrV+Hv7y9SzaRv8eLFKC8vx6hRo8SuiuT89ttvmD17Nn744Qe0aCH9Rz9bbqjRjEYjwsPD8e6776J3794YPXo0XnnlFaxZs0bsqknOvn378NprryE9PR2HDx9GZmYmvvrqK6SlpYldNSKn2Lx5MxYsWICPPvoI4eHhYldHUgwGA8aOHYsFCxagY8eOYlfHLaQfvpFdwsLC4Ovri4KCAovtBQUFiIyMtHlMVFQU5HI5fH19zdu6dOmC/Px86PV6+Pn5ubTO3qox9zolJQVPPPEEJk+eDAC49dZbUVFRgaeffhqvvPIKfHz4d4qzREZG2vxu1Go1W21cZMuWLZg8eTI+/vhjqy5BarqysjIcPHgQR44cwfPPPw+g5o9TQRDQokULfPPNNxgyZIjItXQu/kYkAICfnx969+6N3bt3m7cZjUbs3r0b8fHxNo/p378/Tp8+DaPRaN526tQpREVFMbCpR2PudWVlpVUAYwoqBS4P51Tx8fEW3w0A7Nq1q87vhpomIyMDkyZNQkZGBoYPHy52dSRJrVbjl19+QU5Ojvn17LPPolOnTsjJyUG/fv3ErqLziZzQTB5ky5YtgkKhENavXy8cP35cePrpp4WWLVsK+fn5giAIwhNPPCHMnj3bXP6PP/4QgoKChOeff144efKk8OWXXwrh4eHCwoULxboEr+HovU5NTRWCgoKEjIwM4ffffxe++eYboX379sKoUaPEugSvUVZWJhw5ckQ4cuSIAEBYunSpcOTIEeHcuXOCIAjC7NmzhSeeeMJc/vfffxcCAgKEl156SThx4oSwatUqwdfXV9i5c6dYl+A1HL3XmzZtElq0aCGsWrVKyMvLM7+uXLki1iV4DUfvdW1SHy3F4IYsvP3228LNN98s+Pn5CX379hUOHDhg3jdw4EBhwoQJFuV/+uknoV+/foJCoRDatWsnvPrqq8K1a9fcXGvv5Mi9rq6uFubPny+0b99eUCqVQkxMjDB16lShpKTE/RX3Mnv37hUAWL1M93fChAnCwIEDrY7p0aOH4OfnJ7Rr105Yt26d2+vtjRy91wMHDqy3PNWtMT/XN5J6cCMTBLZpExERkXQw54aIiIgkhcENERERSQqDGyIiIpIUBjdEREQkKQxuiIiISFIY3BAREZGkMLghIiIiSWFwQ0RERJLC4IaIiIgkhcENkZeaOHEiZDJZna8rV66IXUUiIlEwuCHyYvfeey/y8vIsXtu2bRO7WkREomJwQ+TFFAoFIiMjLV4hISFW5bZt24Zu3bpBoVAgNjYWS5Yssdiv0+nw8ssvIyYmBgqFAh06dMB7771n3n/s2DHcd999UKlUiIiIwBNPPIHLly+b9w8aNMjcYuTv748ePXpg586d5v0TJ07EQw89ZPGZ69evR8uWLc3v58+fjx49eti8zk8//RQymcxi22effYZevXpBqVSiXbt2WLBgAa5du1bnvaqrpevGOgDA2bNnG2wJW716Ndq3bw8/Pz906tQJH374ocU5ZDIZPv30U/P7F198EYMGDbL6jJycHPO2pUuX4tZbb0VgYCBiYmIwdepUlJeXW11HbGysVd1u/KxBgwbhxRdfNL8/efIk5HJ5nfeWSIoY3BBJ3KFDhzBq1CiMGTMGv/zyC+bPn4+UlBSsX7/eXGb8+PHIyMjAihUrcOLECbzzzjtQqVQAgCtXrmDIkCHo2bMnDh48iJ07d6KgoACjRo2y+JwpU6YgLy8Px44dQ/fu3TFhwgSXXdMPP/yA8ePHY8aMGTh+/DjeeecdrF+/Hq+++mq9x9Vu6Vq+fLlVGdNawt9++63NlrDt27djxowZ+Otf/4pjx47hmWeewaRJk7B3794mXZOPjw9WrFiB//73v9iwYQP27NmDWbNm2Sz797//3XwNDXnppZegVCqbVDcib9NC7AoQkWstXboUd999N1JSUgAAHTt2xPHjx/Hmm29i4sSJOHXqFD766CPs2rULCQkJAIB27dqZj1+5ciV69uyJ1157zbzt/fffR0xMDE6dOoWOHTsCAAICAhAZGYlr164hPDwcGo3GZde0YMECzJ492xxAtWvXDmlpaZg1axZSU1PrPM7U0mViq47V1dUAUGdL2OLFizFx4kRMnToVADBz5kwcOHAAixcvxuDBgxt9TTe2tsTGxmLhwoV49tlnkZ6eblFOp9MhJCTE4jrqsnfvXvz000+YPHlyk4MvIm/ClhsiiTtx4gT69+9vsa1///747bffYDAYkJOTA19fXwwcONDm8UePHsXevXuhUqnMr86dOwMAzpw5Yy6Xnp4OlUoFf39/fPjhh9iwYYPFeb788kuLczz77LNWn/XLL79ApVJBo9GgS5cueP311+us09///neL85lajiorKx26P7VptVoAQGBgoM39dd3PEydONOlzv/32W9x9991o3bo1goKC8MQTT6CoqMjqeoqLi6FWqxs8nyAI+Otf/4rU1FSXBppEnogtN0TNnL+/f737y8vLMWLECLzxxhtW+6Kiosz/fvzxx/HKK6+gqqoKGzZswGOPPYbjx4+bH8SDBw/G6tWrzeUzMzMtWoMAoFOnTvj8889hMBhw4MABTJkyBR06dECLFpa/qsrLy7FgwQKMHDnSqk5N7YK5ePEifHx87GoZcZazZ8/i/vvvx3PPPYdXX30VISEh+PHHH/HUU09Br9cjICAAAPDnn39Cr9ejbdu2DZ7zgw8+QEVFBZ599tkGu+uIpIbBDZHEdenSBfv377fYtn//fnTs2BG+vr649dZbYTQa8d1335m7pW7Uq1cvbNu2DbGxsVZBxo00Gg06dOgAAEhNTcXixYuRnZ1tPmdgYKB5PwCEh4dbncPPz89cplOnTli5ciVycnLQp08fqzqdPHnS4nzO8vPPP6Nz5851Bkmm+3ljTtH+/fvRtWvXRn/moUOHYDQasWTJEvj41DSof/TRR1blvvvuO/j7+1vdj9oqKyvxyiuvYOXKlZDL5Y2uF5G3YnBDJHF//etfcfvttyMtLQ2jR49GVlYWVq5cac7liI2NxYQJE/Dkk09ixYoViIuLw7lz51BYWIhRo0Zh2rRpWLt2LZKSkjBr1iyEhITg9OnT2LJlC/75z3/C19cXQM0DNT8/HzqdDhs2bECLFi0cDj4EQUBVVRUMBgP+/e9/4/jx4/jrX/9qVW7evHm4//77cfPNN+PRRx+Fj48Pjh49imPHjmHhwoWNuk96vR5bt27F0qVLsWDBgjrLvfTSSxg1ahR69uyJhIQEfPHFF8jMzMS3335rUa66uhpVVVUAAIPBAKPRaH6v0+nMnykIAjp06IDq6mq8/fbbGDFiBPbv3481a9ZYnO/MmTN4/fXX8eCDD1rNYXTlyhXo9Xr4+fkBADZv3ozevXtbjVAjajYEIvJKEyZMEB588EGr7Xv37hUACCUlJeZtn3zyidC1a1dBLpcLN998s/Dmm29aHHP16lUhOTlZiIqKEvz8/IQOHToI77//vnn/qVOnhIcfflho2bKl4O/vL3Tu3Fl48cUXBaPRKAiCIAwcOFAAIAAQ/Pz8hG7duglbt26tt67r1q0TNBqN+X1qaqr5HD4+PkJMTIyQkpIiCIIgbN++Xaj962rnzp3CnXfeKfj7+wtqtVro27ev8O677zp0v26sw8GDB4V27doJixYtEgwGQ733Mz09XWjXrp0gl8uFjh07Ch988IHFeU3XYc8rNzdXEARBWLp0qRAVFSX4+/sLiYmJwgcffGDxuW3atKn3PHv37jV/FzKZTPj5558t7m1cXFyd94ZIamSC8L9xj0RE5FY9evTAp59+itjY2AbLxsbGYt++fTbLPvTQQ1Zz6RA1ZxwtRUQkEoVCYTU5YV1atWpl7gKsLTg42NwlRUQAW26IiIhIUthyQ0RERJLC4IaIiIgkhcENERERSQqDGyIiIpIUBjdEREQkKQxuiIiISFIY3BAREZGkMLghIiIiSWFwQ0RERJLy/9/L5N2VvJmaAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "plt.figure()\n",
    "\n",
    "plt.scatter(\n",
    "    features[\"Динамика_Посевных_Площадей\"],\n",
    "    features[\"Динамика_Величина_КРС\"],\n",
    ")\n",
    "\n",
    "plt.scatter(\n",
    "    features[pred_iso==1][\"Динамика_Посевных_Площадей\"],\n",
    "    features[pred_iso==1][\"Динамика_Величина_КРС\"],\n",
    "    marker=\"x\"\n",
    ")\n",
    "\n",
    "plt.title(\"Аномалии — Isolation Forest\")\n",
    "plt.xlabel(\"Посевные площади\")\n",
    "plt.ylabel(\"КРС\")\n",
    "\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "ccd62a95",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "plt.figure()\n",
    "\n",
    "plt.scatter(\n",
    "    features[\"Динамика_Посевных_Площадей\"],\n",
    "    features[\"Динамика_Величина_КРС\"],\n",
    ")\n",
    "\n",
    "plt.scatter(\n",
    "    features[pred_nn==1][\"Динамика_Посевных_Площадей\"],\n",
    "    features[pred_nn==1][\"Динамика_Величина_КРС\"],\n",
    "    marker=\"x\"\n",
    ")\n",
    "\n",
    "plt.title(\"Аномалии — Autoencoder\")\n",
    "plt.xlabel(\"Посевные площади\")\n",
    "plt.ylabel(\"КРС\")\n",
    "\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "05286213",
   "metadata": {},
   "source": [
    "\n",
    "## 8. Вывод\n",
    "\n",
    "В работе были использованы два метода поиска аномалий:\n",
    "\n",
    "1. **Isolation Forest** — классический алгоритм машинного обучения.\n",
    "2. **Autoencoder** — нейросетевая модель.\n",
    "\n",
    "После расчёта метрик **Precision, Recall и F1-score** можно сравнить методы.\n",
    "\n",
    "Обычно:\n",
    "- Isolation Forest показывает стабильные результаты на небольших датасетах.\n",
    "- Нейросети лучше работают на больших и сложных данных.\n",
    "\n",
    "В данном датасете (небольшого размера) **Isolation Forest часто показывает более устойчивые результаты**, поэтому его можно считать **более адекватным методом для данной задачи**.\n"
   ]
  }
 ],
 "metadata": {
  "language_info": {
   "name": "python"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}