notebooks/LSTM_AE.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Pipeline for the anomaly detection on the SKAB using LSTM-Autoencoder"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Details regarding the SKAB one can find in the [SKAB repository](https://github.com/waico/SKAB)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# libraries importing\n",
    "import sys\n",
    "import warnings\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "sys.path.append(\"..\")\n",
    "from core.LSTM_AE import LSTM_AE\n",
    "from core.metrics import chp_score\n",
    "from core.utils import create_sequences, load_preprocess_skab, plot_results\n",
    "\n",
    "warnings.filterwarnings(\"ignore\", category=UserWarning)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Xy_traintest_list = load_preprocess_skab()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Method"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Method initialization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# hyperparameters selection\n",
    "EPOCHS = 100\n",
    "BATCH_SIZE = 32\n",
    "VAL_SPLIT = 0.1\n",
    "N_STEPS = 10\n",
    "Q = 0.99  # quantile for upper control limit (UCL) selection\n",
    "PARAM = [EPOCHS, BATCH_SIZE, VAL_SPLIT]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = LSTM_AE(PARAM)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Method fitting and applying"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# inference\n",
    "predicted_outlier, predicted_cp = [], []\n",
    "true_outlier, true_cp = [], []\n",
    "for X_train, X_test, y_train, y_test in Xy_traintest_list:\n",
    "    # scaler init and fitting\n",
    "    StSc = StandardScaler()\n",
    "    StSc.fit(X_train)\n",
    "\n",
    "    # convert into input/output\n",
    "    X = create_sequences(StSc.transform(X_train), N_STEPS)\n",
    "\n",
    "    # model fitting\n",
    "    # history, model = arch(X)\n",
    "    model.fit(X)\n",
    "\n",
    "    # results predicting\n",
    "    residuals = pd.Series(\n",
    "        np.sum(np.mean(np.abs(X - model.predict(X)), axis=1), axis=1)\n",
    "    )\n",
    "    UCL = residuals.quantile(Q) * 3 / 2\n",
    "\n",
    "    # results predicting\n",
    "    X = create_sequences(StSc.transform(X_test), N_STEPS)\n",
    "    cnn_residuals = pd.Series(\n",
    "        np.sum(np.mean(np.abs(X - model.predict(X)), axis=1), axis=1)\n",
    "    )\n",
    "\n",
    "    # data i is an anomaly if samples [(i - timesteps + 1) to (i)] are anomalies\n",
    "    anomalous_data = cnn_residuals > UCL\n",
    "    anomalous_data_indices = []\n",
    "    for data_idx in range(N_STEPS - 1, len(X) - N_STEPS + 1):\n",
    "        if np.all(anomalous_data[data_idx - N_STEPS + 1 : data_idx]):\n",
    "            anomalous_data_indices.append(data_idx)\n",
    "\n",
    "    prediction = pd.Series(data=0, index=X_test.index)\n",
    "    prediction.iloc[anomalous_data_indices] = 1\n",
    "\n",
    "    # predicted outliers saving\n",
    "    predicted_outlier.append(prediction)\n",
    "\n",
    "    # predicted CPs saving\n",
    "    prediction_cp = abs(prediction.diff())\n",
    "    prediction_cp[0] = prediction[0]\n",
    "    predicted_cp.append(prediction_cp)\n",
    "\n",
    "    true_outlier.append(y_test[\"anomaly\"])\n",
    "    true_cp.append(y_test[\"changepoint\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Results visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA+MAAAIQCAYAAADq9b/TAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8fJSN1AAAACXBIWXMAAA9hAAAPYQGoP6dpAACV4UlEQVR4nO3deXxU1f3/8fdMVrYASghbIIAL4AIKkqK1gEWhVSxFK64sblVBkWgFXFi0ErWF4o5LEaz4K4qKVvziEkFFsRYpaqsgIAjKIlRJIEACM/f3RzI3c2fuhEwyNzOZ+3o+HlFy5y5nTubOnM98zuIxDMMQAAAAAACoN954FwAAAAAAALchGAcAAAAAoJ4RjAMAAAAAUM8IxgEAAAAAqGcE4wAAAAAA1DOCcQAAAAAA6hnBOAAAAAAA9YxgHAAAAACAekYwDgAAAABAPSMYBwAgwSxfvlwej0fLly939DqjR49WXl6eo9dwyrx58+TxeLR58+Z4FwUAgFohGAcAAFF77LHHNG/ePMevM2PGDC1evNjx6wAAUN8IxgEAQNTiHYxfccUVOnDggDp16uR4GQAAcEJqvAsAAAAQrZSUFKWkpMS7GAAA1BqZcQBAg/Ltt9/qhhtu0PHHH69GjRrp6KOP1u9+97uwscOBMcUffvihCgoKlJ2drSZNmui3v/2tdu3aFXbexx57TCeccIIyMjLUrl07jR07Vnv27LHsM2DAAJ144on6/PPP1b9/fzVu3FjHHHOMFi1aJEl67733lJ+fr0aNGun444/XO++8U6uyh5o6darS0tJsy33ttdeqRYsWOnjwYLXnWLx4sU488URlZmbqxBNP1CuvvGK7n9/v1+zZs3XCCScoMzNTOTk5+v3vf6+ffvrJ3CcvL0///e9/9d5778nj8cjj8WjAgAHm43v27NHNN9+s3NxcZWRk6JhjjtH9998vv98fdq0HH3xQJ510kjIzM5Wdna0hQ4Zo1apVkiSPx6PS0lLNnz/fvM7o0aMlRR4zHs3f8csvv9TAgQPVuHFjtW/fXg888EC1dQgAQCwRjAMAGpR//etf+uijj3TxxRfroYce0nXXXaeioiINGDBA+/fvD9v/xhtv1GeffaapU6fq+uuv1z/+8Q+NGzfOss+0adM0duxYtWvXTjNnztQFF1ygJ554Quecc44OHTpk2fenn37Seeedp/z8fD3wwAPKyMjQxRdfrIULF+riiy/Wr3/9a913330qLS3VhRdeqL1799a67AFXXHGFDh8+rIULF1q2l5eXa9GiRbrggguUmZkZ8fi33npLF1xwgTwejwoLCzVs2DCNGTPGDHqD/f73v9cf/vAHnXHGGXrwwQc1ZswYLViwQIMHDzbrYvbs2erQoYO6deumv/3tb/rb3/6mO+64Q5K0f/9+9e/fX88995xGjhyphx56SGeccYYmT56sgoICy7WuuuoqM2i///77NWnSJGVmZurjjz+WJP3tb39TRkaGzjzzTPM6v//97yM+z2j/jkOGDFHPnj01c+ZMdevWTRMnTtT//d//RTw/AAAxZQAA0IDs378/bNvKlSsNScazzz5rbnvmmWcMScagQYMMv99vbp8wYYKRkpJi7NmzxzAMw/jhhx+M9PR045xzzjF8Pp+53yOPPGJIMubOnWtu69+/vyHJeP75581ta9euNSQZXq/X+Pjjj83tb775piHJeOaZZ6Iu+7JlywxJxrJly8xt/fr1M/Lz8y3Hvvzyy2H72enVq5fRtm1b8zkbhmG89dZbhiSjU6dO5rYPPvjAkGQsWLDAcvzSpUvDtp9wwglG//79w651zz33GE2aNDG+/vpry/ZJkyYZKSkpxpYtWwzDMIx3333XkGTcdNNNYecI/ns1adLEGDVqVNg+gb/vpk2bDMOo3d8xuM7LysqMNm3aGBdccEHYtQAAcAKZcQBAg9KoUSPz34cOHdL//vc/HXPMMWrRooVWr14dtv+1114rj8dj/n7mmWfK5/Pp22+/lSS98847Ki8v18033yyvt+pj8ZprrlFWVpaWLFliOV/Tpk118cUXm78ff/zxatGihbp37678/Hxze+Df33zzTa3LHmzkyJH65z//qY0bN5rbFixYoNzcXPXv3z/icdu3b9eaNWs0atQoNW/e3Nx+9tlnq0ePHpZ9X3zxRTVv3lxnn322du/ebf707t1bTZs21bJly6otY+AcZ555plq2bGk5x6BBg+Tz+fT+++9Lkl566SV5PB5NnTo17BzBf6+aqs3f8fLLLzd/T09PV9++fS1/LwAAnEQwDgBoUA4cOKApU6aY45FbtWql7Oxs7dmzR8XFxWH7d+zY0fJ7y5YtJckcAx0Iyo8//njLfunp6erSpYv5eECHDh3CgsXmzZsrNzc3bFvwdWpT9mAjRoxQRkaGFixYIEkqLi7W66+/rssuu6za4DVQ/mOPPTbssdDnvH79ehUXF6t169bKzs62/Ozbt08//PBDtWUMnGPp0qVhxw8aNEiSzHNs3LhR7dq101FHHXXEc9ZELP6OLVu2tPy9AABwErOpAwAalBtvvFHPPPOMbr75ZvXr10/NmzeXx+PRxRdfHDZBmKSIM24bhlGr60c6X02uE23Zg7Vs2VLnnXeeFixYoClTpmjRokUqKyuzZHfryu/3q3Xr1mbAHyo7O7tG5zj77LN122232T5+3HHH1amMsRLr1wUAANEiGAcANCiLFi3SqFGjNHPmTHPbwYMHw2bMrqnAOtXr1q1Tly5dzO3l5eXatGmTmdGNhbqWfeTIkfrNb36jf/3rX1qwYIFOOeUUnXDCCdUeE3h+69evD3ts3bp1lt+7du2qd955R2eccYalS72dSNn4rl27at++fUest65du+rNN9/Ujz/+WG12vKZd1uvz7wgAQCzQTR0A0KCkpKSEZS8ffvhh+Xy+Wp1v0KBBSk9P10MPPWQ571//+lcVFxfr3HPPrVN5g9W17L/61a/UqlUr3X///XrvvfdqlBVv27atevXqpfnz51u6wr/99tv68ssvLftedNFF8vl8uueee8LOc/jwYcuXBk2aNLH9EuGiiy7SypUr9eabb4Y9tmfPHh0+fFiSdMEFF8gwDE2fPj1sv+A6inSdUPX5dwQAIBbIjAMAGpTzzjtPf/vb39S8eXP16NFDK1eu1DvvvKOjjz66VufLzs7W5MmTNX36dA0ZMkTnn3++1q1bp8cee0ynnXZaTLuB17XsaWlpuvjii/XII48oJSVFl1xySY2OKyws1Lnnnquf//znuvLKK/Xjjz/q4Ycf1gknnKB9+/aZ+/Xv31+///3vVVhYqDVr1uicc85RWlqa1q9frxdffFEPPvigLrzwQklS79699fjjj+uPf/yjjjnmGLVu3VpnnXWW/vCHP+i1117Teeedp9GjR6t3794qLS3VF198oUWLFmnz5s1q1aqVBg4cqCuuuEIPPfSQ1q9fryFDhsjv9+uDDz7QwIEDzeXnevfurXfeeUezZs1Su3bt1LlzZ8tEeQH1+XcEACAWCMYBAA3Kgw8+qJSUFC1YsEAHDx7UGWecoXfeeUeDBw+u9TmnTZum7OxsPfLII5owYYKOOuooXXvttZoxY4bS0tISquwjR47UI488ol/+8pdq27ZtjY4ZMmSIXnzxRd15552aPHmyunbtqmeeeUavvvqqli9fbtl3zpw56t27t5544gndfvvtSk1NVV5eni6//HKdccYZ5n5TpkzRt99+qwceeEB79+5V//79ddZZZ6lx48Z67733NGPGDL344ot69tlnlZWVpeOOO07Tp0+3zOj+zDPP6OSTT9Zf//pX/eEPf1Dz5s3Vp08fnX766eY+s2bN0rXXXqs777xTBw4c0KhRo2yDcan+/o4AAMSCx2CmEgAAGozPPvtMvXr10rPPPqsrrrgi3sUBAAC1xJhxAAAakKeeekpNmzbV8OHD410UAABQB3RTBwCgAfjHP/6hL7/8Uk8++aTGjRunJk2axLtIAACgDuimDgBAA5CXl6edO3dq8ODB+tvf/qZmzZrFu0gAAKAOCMYBAAAAAKhnjBkHAAAAAKCeEYwDAAAAAFDPGsQEbn6/X9u2bVOzZs3k8XjiXRwAAAAAQJIzDEN79+5Vu3bt5PXGPo/dIILxbdu2KTc3N97FAAAAAAC4zNatW9WhQ4eYn7dBBOOBGWO3bt2qrKysOJcGAAAAAJDsSkpKlJub69gKJg0iGA90Tc/KyiIYBwAAAADUG6eGSjOBGwAAAAAA9YxgHAAAAACAehZ1MP7+++9r6NChateunTwejxYvXnzEY5YvX65TTz1VGRkZOuaYYzRv3rxaFBUAAAAAgOQQ9Zjx0tJS9ezZU1deeaWGDx9+xP03bdqkc889V9ddd50WLFigoqIiXX311Wrbtq0GDx5cq0IDaDg27S7Vk+9v1Icb/qef9pfLI6lpRqoapafosM+QISktxaMD5T7tLTssv9+wPU+K12Med6R9qztu/yGfGqelHPE8XK9m19tbdtjyN410rlhez+83jni+WF+vruXhekc+LnjfaN4bEu16Tt0PiXQv1Pax2l6vLs8/Ea932GcoLdWrvnlH6ZpfdFHnVk2OeDwSl8/n06FDh+JdDNRSWlqaUlJS4nZ9j2EYR34HiXSwx6NXXnlFw4YNi7jPxIkTtWTJEv3nP/8xt1188cXas2ePli5dWqPrlJSUqHnz5iouLmYCN6ABeWHVVk1c9Llq/SaThG5OXaS+nq8s2wx51NHzg7I8pea2EqOJDEnNg7aFKjaayCOZxxUbTbRNrdROu83j7M4TelxtrydJ//XnSZJO8G6OeK6aXK/YaKLvjGzL9T4xumv24Qsjlqc6dvXcwbMr6ucXXM5Ij28xWsfk7+e260XaN/BvSea+DeV6sbofavrcjvS3ieW9HjhOUlSP/defp5X+E9TP+19zW13u7WTjkfTAhSfrd31YwrehMQxDO3bs0J49e+JdFNRRixYt1KZNG9tJ2pyOQx2fTX3lypUaNGiQZdvgwYN18803O31pAHG0aXcpgbgNn+HV6alfHXG/5p4DUe/T3HNALf371Mx7sNrz1HRbTfY5PSX8udiVqyb7dNRuy7aVh044YpkiqWk9V1em0G2RHs8NKXekfble5ONqUo6GcL1Y3Q81ebwmf5v6qMsjPXZ6yldh9VKXezvZGJImvvS5Tss7SnlkyBuUQCDeunVrNW7c2LHZtuEcwzC0f/9+/fDDD5Kktm3b1nsZHA/Gd+zYoZycHMu2nJwclZSU6MCBA2rUqFHYMWVlZSorKzN/LykpcbqYAGLshVVb412EhPSwr2J4zy1pixw5f3Ag3pDNPHShWVe18bBvuI7zfKehqR/HsFQA6qqu93YyMgxp4aqtmjikW7yLghry+XxmIH700UfHuziog0As+sMPP6h169b13mU9IWdTLywsVPPmzc2f3Fy67gANzXc/HSArHsHDvuGKNECotgOHQo+r6Xlidb2anqsm+5QbKTFprH9tdKj1sUcqp9N/P7ddL9K+Dfl6sbofanpMbf+mtb1ebR4rM1IJxG0YqvjMRMMRGCPeuHHjOJcEsRD4O8Zj7L/jwXibNm20c+dOy7adO3cqKyvLNisuSZMnT1ZxcbH5s3UrGTagoenQspHosGXvxpSXFak3W217uYUeV9PzxOp6NT1XTfZJ9/h0Y8rL0RcqxM+8X9b62COV0+m/n9uuF2nfhny9WN0PNT2mtn/T2l6vNo9leA7H5N5ONh5VfGai4aFrenKI59/R8WC8X79+Kioqsmx7++231a9fv4jHZGRkKCsry/IDoGG5iMlobN2Y8rJjXdQlaY8/Ob6lvyVtUZ0a7TemvKwzUmofjANwRl3v7WTk8Ugj+MwEXCnqMeP79u3Thg0bzN83bdqkNWvW6KijjlLHjh01efJkff/993r22WclSdddd50eeeQR3Xbbbbryyiv17rvv6oUXXtCSJUti9ywAJJzOrZrogQtP1m1M4maR4vHrI1939fZ+rQyPT5/5O2uf0Si2s6n7G+Zs6m09PyrN49e/fV11QOlK8fgjluVIUjx+bfW3Uq53t7b7W2qT0YbZ1BPsepH2ZTb1mj+3hjCb+jajlbp7t2qfkanP/Z0lqU73drLxSLr/gpOZvA1wqaiD8VWrVmngwIHm7wUFBZKkUaNGad68edq+fbu2bNliPt65c2ctWbJEEyZM0IMPPqgOHTro6aefZo1xwAV+1ydXp+UdpfMfWaGSg4fVON2rVK9XTTNS1Tg9VYd8fhmS0lO82l9+WHvLDsuIsF6rt3KN1sbpqUfct7rjAutwH+k8Tl3vr2UX62m/oWX6vbK1R3frOq3z5EW+3hEuWePnZ1R/XG2vF7yucnXnqsn1PtDVaqm9mua5XhuVK3mkrEz7447095nrvVjNU7wa41ukD1N/pun+MbV+fsHHRXo8lvXpputF2jfie0MDuF6s7oeaPrfa/k1rXZdRPtaz7F96XIXaoja61jOt4uQeqWl6za9Xq+d/BPG6XpnPr//tK1fTjBSd37O9rv1FFwJx1JsjdceeOnWqpk2bVj+FgaRaBOMDBgxQdUuTz5s3z/aYf//739FeCkASyGvVRK2zMlVycJ/+Oqqv+nVl1lFJ0p/SpVLppRt+LrU5Md6lSQwPZEj79+rVsWdIOT3qfr6iT6QPpAv7dNSFv+YLYCAuNhjSc1KPts30n+uGxLs0cTfvw02a9o8v1f/41pox/KR4FwcJYNPuUr2waqu+++mAOrRspIv65KqzQ1/QbN++3fz3woULNWXKFK1bt87c1rRpU/PfhmHI5/MpNdXxxbdcLSFnUweQXAJf4DHPSbDKLzWplCqe4M7CsUAdA/FXef8xXklSUGaS+oAqloH95czlevL9b7Tk82168v1v9MuZy/WiQ8vDtmnTxvxp3ry5PB6P+fvatWvVrFkz/d///Z969+6tjIwMrVixQqNHj9awYcMs57n55ps1YMAA83e/36/CwkJ17txZjRo1Us+ePbVokXPz4yQTvuoA4LhAZxovQVEVo3LMpIfvRE2BujBiNJ6UOgbiL9b3dQPnrfwY9Nd2rTkkLMMwdOCQr8b7b/5fqSa99Ln8hqoaSpX/n/jS5zqxfZY6HV2zDHmjtJSYzQg+adIk/fnPf1aXLl3UsmXLGh1TWFio5557TnPmzNGxxx6r999/X5dffrmys7PVv3//mJQrWRGMA3CcObEPsXgVsyFGpVQJZIxi1EiljoH4i3mPlwausj6IxZPPgUM+9ZjyZkzO5TekXz24osb7f3n3YDVOj01Yd/fdd+vss8+u8f5lZWWaMWOG3nnnHXO1rC5dumjFihV64oknCMaPgGAcgOMCGQAvMVEVsrbhHMuM88ID4obMuAWZcSS6Pn36RLX/hg0btH///rAAvry8XKecckosi5aUCMYBOK6qzUFQVIXxzGEYMw4koRj3eGngPJX1QW0kn0ZpKfry7ppPFjrr7a/1zIrN8tncGykej8b8PE8FZx9X42vHSpMm1q7xXq83bPLuQ4cOmf/et2+fJGnJkiVq3769Zb+MjIyYlStZEYwDcByZcRuBDzYy41VinhmnjoG4IzNu4TW/myAcTzYejyeqruKX5XfS3BWbbB8zZOjy/E4x63peF9nZ2frPf/5j2bZmzRqlpaVJknr06KGMjAxt2bKFLum1QAsFgOPMmIgMZRUaYjZiPMswY8aB+GPMuIWHjgKo1LlVE91/wcnyeqQUr8fy//svODlh1p8/66yztGrVKj377LNav369pk6dagnOmzVrpltvvVUTJkzQ/PnztXHjRq1evVoPP/yw5s+fH8eSNwzx/7oFQNIzlzaLczkSCuOZw5mtVMaMA8kjxvd1Axfops6YcUjS7/rk6rS8o7QwaJ3xEX1yEyYQl6TBgwfrrrvu0m233aaDBw/qyiuv1MiRI/XFF1+Y+9xzzz3Kzs5WYWGhvvnmG7Vo0UKnnnqqbr/99jiWvGEgGAfguECTg6XNgtGFOoxjY8apYyBuzG7qBJ9S0HeO8S0GEkheqyaaOKRbvV939OjRGj16tPn7gAEDIg6fmD59uqZPnx7xXB6PR+PHj9f48eNjXcykRwsFgOMM5tEKRxdqGyxtBiQduqlbeFjaDEAQgnEAjgt0xyMYD8LSZuEcW9qMOgbihgncLFjaDEAwWigAHFeVnyQar0J3gTAsbQYkIfplB+PtCEAwgnEAjguMQfLyjlOFrG04MuNA8on1xIwNXGDuFDLjACSCcQD1wBwzTma8CuOZbTBmHEg6jBm3RSwOQCIYB1APAhkALzFRFbK24ciMA8mHMeMWZMYBBKOFAsBxZn6SYDwI45nDODZmPEanA1ALMe7x0sB5qA4AQQjGATjO76e7sIWlFUadmJzKjFPHQPwwZtwiMFyLYByARDAOoB4E2hx0U68U3AqjC3WQWI8ZD5yWOgbixrz/iD6lqs9Bg/oAIIJxAPXB7JFNNF4hOBinTkxmVbC0GZA86JcdjG7qcIvRo0dr2LBh5u8DBgzQzTffXO/lWL58uTwej/bs2VPv164JgnEAjmMCtxDB3TUJFKuY3dRjlRlnAjcg7pjAzcLDBG6Is9GjR8vj8cjj8Sg9PV3HHHOM7r77bh0+fNjR67788su65557arRvogfQsZQa7wIASH5VI8YJPCUxZjwiljYDkg5Lm1lQG0gEQ4YM0TPPPKOysjK98cYbGjt2rNLS0jR58mTLfuXl5UpPT4/JNY866qiYnCfZkC4A4LhABoAkcCVLZpy3YRNLmwHJh8y4RdXSZnEuCBLDskLpvQes2957oGK7gzIyMtSmTRt16tRJ119/vQYNGqTXXnvN7Fp+7733ql27djr++OMlSVu3btVFF12kFi1a6KijjtJvfvMbbd682Tyfz+dTQUGBWrRooaOPPlq33XabjJAv1kO7qZeVlWnixInKzc1VRkaGjjnmGP31r3/V5s2bNXDgQElSy5Yt5fF4NHr0aEmS3+9XYWGhOnfurEaNGqlnz55atGiR5TpvvPGGjjvuODVq1EgDBw60lDMR0UIB4DiDobshGDNuy7GlzahjIH7MGcug4Lc5KiTpGIZUXhrdj+GTlt0rvfvHit/f/WPF74YvyvPU7fXUqFEjlZeXS5KKioq0bt06vf3223r99dd16NAhDR48WM2aNdMHH3ygDz/8UE2bNtWQIUPMY2bOnKl58+Zp7ty5WrFihX788Ue98sor1V5z5MiR+n//7//poYce0ldffaUnnnhCTZs2VW5url566SVJ0rp167R9+3Y9+OCDkqTCwkI9++yzmjNnjv773/9qwoQJuvzyy/Xee+9JqvjSYPjw4Ro6dKjWrFmjq6++WpMmTapT3TiNbuoAHBf4jPASFFUgM26PzDiQfFjazILMeBI7tF+a0a52x77/p4qfSL8fye3bpPQmUV/WMAwVFRXpzTff1I033qhdu3apSZMmevrpp83u6c8995z8fr+efvppc86DZ555Ri1atNDy5ct1zjnnaPbs2Zo8ebKGDx8uSZozZ47efPPNiNf9+uuv9cILL+jtt9/WoEGDJEldunQxHw90aW/durVatGghqSKTPmPGDL3zzjvq16+fecyKFSv0xBNPqH///nr88cfVtWtXzZw5U5J0/PHH64svvtD9998fdd3UF4JxAI4LLOFCLF6JMeMRMGYcSDqMGbdiaTMkgNdff11NmzbVoUOH5Pf7demll2ratGkaO3asTjrpJMs48c8++0wbNmxQs2bNLOc4ePCgNm7cqOLiYm3fvl35+fnmY6mpqerTp09YV/WANWvWKCUlRf37969xmTds2KD9+/fr7LPPtmwvLy/XKaecIkn66quvLOWQZAbuiYpgHIDjzG7qBEWV6KZuK+brEdNNHYg/1vIKRi/1JJbWuCJDHa0Vf6nIgqekS75y6Rd/kH4+IfprR2HgwIF6/PHHlZ6ernbt2ik1tSokbNLEmmHft2+fevfurQULFoSdJzs7O7pyVmrUqFHUx+zbt0+StGTJErVv397yWEZGRq3KkQgIxgE4jqXNQtBN3V6su7PSTR2IPyZws6CbehLzeKLvKv7eAxWB+MA7pP63VU7edm9FYN7/NmfKqYqA+5hjjqnRvqeeeqoWLlyo1q1bKysry3aftm3b6p///Kd+8YtfSJIOHz6sTz/9VKeeeqrt/ieddJL8fr/ee+89s5t6sEBm3ufzmdt69OihjIwMbdmyJWJGvXv37nrttdcs2z7++OMjP8k4ooUCwHFmm4NgvALd1CNwqps6gLihm7qF+Z0j70+QJL+vKhCXKv4/8I6K7QnisssuU6tWrfSb3/xGH3zwgTZt2qTly5frpptu0nfffSdJGj9+vO677z4tXrxYa9eu1Q033FDtGuF5eXkaNWqUrrzySi1evNg85wsvvCBJ6tSpkzwej15//XXt2rVL+/btU7NmzXTrrbdqwoQJmj9/vjZu3KjVq1fr4Ycf1vz58yVJ1113ndavX68//OEPWrdunZ5//nnNmzfP6Sqqk1oF448++qjy8vKUmZmp/Px8ffLJJ9XuP3v2bB1//PFq1KiRcnNzNWHCBB08eLBWBQbQ8DCBWwgy4/aYwA1IPmTGLQKfg8TikCQNnByeAe9/W8X2BNG4cWO9//776tixo4YPH67u3bvrqquu0sGDB81M+S233KIrrrhCo0aNUr9+/dSsWTP99re/rfa8jz/+uC688ELdcMMN6tatm6655hqVlpZKktq3b6/p06dr0qRJysnJ0bhx4yRJ99xzj+666y4VFhaqe/fuGjJkiJYsWaLOnTtLkjp27KiXXnpJixcvVs+ePTVnzhzNmDHDwdqpO48R5VdzCxcu1MiRIzVnzhzl5+dr9uzZevHFF7Vu3Tq1bt06bP/nn39eV155pebOnavTTz9dX3/9tUaPHq2LL75Ys2bNqtE1S0pK1Lx5cxUXF0fsHgEgMRmGoc6T35AkfXrnIB3dtOGO64mZ/T9KD1R8cGjKj5I3Jb7lSRTzzpM2fyBdOFc68YK6n+/vl0lrX5fO+4vU58q6nw9A9Iq/l/7SQ/KmSVN2x7s0cffRht269Ol/6ricpnprQs0nr0JiOXjwoDZt2qTOnTsrMzMz3sVBHVX393Q6Do06XTBr1ixdc801GjNmjHr06KE5c+aocePGmjt3ru3+H330kc444wxdeumlysvL0znnnKNLLrnkiNl0AMkheFwcmfFKZMbteWLdTZ3MOBB3ZMYtPIwZBxAkqhZKeXm5Pv30U8tAe6/Xq0GDBmnlypW2x5x++un69NNPzeD7m2++0RtvvKFf//rXEa9TVlamkpISyw+Ahim48w2xeCWD2dTtsbQZkHQYM27BmHEAwaKaTX337t3y+XzKycmxbM/JydHatWttj7n00ku1e/du/fznP5dhGDp8+LCuu+463X777RGvU1hYqOnTp0dTNAAJym+JOwmKJAVliKgPC8aMA8mHzLgFY8YBBHO8hbJ8+XLNmDFDjz32mFavXq2XX35ZS5Ys0T333BPxmMmTJ6u4uNj82bp1q9PFBOAQQ2TGw7H+ta2YZ9CoZyD+uP+CmZnx+BYDQIKIKjPeqlUrpaSkaOfOnZbtO3fuVJs2bWyPueuuu3TFFVfo6quvllSxrlxpaamuvfZa3XHHHfJ6w78PyMjIaNCLtwOoYumRHb9iJBa6T9szM2h0UweSRvCXYYbh+i/HzK8cSY0DUJSZ8fT0dPXu3VtFRUXmNr/fr6KiIvXr18/2mP3794cF3CkpFTMH80YEJD+DCdzC0X06gkDKiG7qQNIIvv/oqs4EbknG7+c1nQzi+XeMKjMuSQUFBRo1apT69Omjvn37avbs2SotLdWYMWMkSSNHjlT79u1VWFgoSRo6dKhmzZqlU045Rfn5+dqwYYPuuusuDR061AzKASQvuqnbofu0LbPRTjd1ICmRhAnqpk5dNGTp6enyer3atm2bsrOzlZ6ezrw4DZBhGCovL9euXbvk9XqVnp5e72WIOhgfMWKEdu3apSlTpmjHjh3q1auXli5dak7qtmXLFksm/M4775TH49Gdd96p77//XtnZ2Ro6dKjuvffe2D0LAAmLpc1skLG15yEzDiQdMuMWgc9BEqoNm9frVefOnbV9+3Zt27Yt3sVBHTVu3FgdO3a0HT7ttKiDcUkaN26cxo0bZ/vY8uXLrRdITdXUqVM1derU2lwKQAPHcBQbjGWOgKXNgKRj+RKWzwPejZJHenq6OnbsqMOHD8vn88W7OKillJQUpaamxq1nQ62CcQCoKTLjNsjY2nNsaTNed0DckBm3MDPjfFGdFDwej9LS0pSWlhbvoqCBoiUIwFmWdcbjV4zEwlhmWzFf2iz0vADqX8hs6i7niXEHIAANG8E4AEcFf/tPZrxSoE7IjFsxZhxIPmTGLQJvc2TGAUgE4wAcFtzcIBSvxFjmCBgzDiQdxoxbeCrfj6gJABLBOACHBU/gRmI8IJAZj28pEk7MM9gMBwDij27qweimDiAYwTgAR/ktY8YJiiTRfToSuqkDyYdu6haB4VqsNAJAIhgH4DCjMjtJHB6E7tP2zNnU6aYOJA26qVuY3znGtxgAEgTBOABHBeIhJm8LQsY2AjLjQNKxZMYJQb1M4AYgCC0UAI4yGB5tg7HMtmK+tBn1DMQfY8atAt3U41wMAAmBYByAowLf/pMZD0LG1p7ZTZ3MOJA0gt/7GTNOZhyABS0UAI4ymxvE4lUYyxwBS5sBSYcx4xYeBo0DCEIwDsBRfn8gMx7ngiQSMrb2yIwDySnW93YDRmYcQDBaKADqhYfsZBDGMttybMx4jE4HoJZYXDsg8FlITQCQCMYBOMwg7gxH92l7nlh3UzdPHJvzAaidmH/R1nDF+m0OQMNGMA7AUUzgZsP8hoK3YCuWNgOSEt3UTR66qQMIQgsFgKPIAduh+7QtM2hmaTMguZAODghM4EZNAJAIxgE4LPDtP/FQEDK29mLeTZ16BhICmXGT13ybIxwHQDAOwGFVY8aJxk2MGY+Apc2ApMSYcZM5gRtVAUAE4wAcZhgsbRaGjK09ljYDkhOZcRNLmwEIRgsFgKPM3CSZ8SCMZbbl2NJm1DMQX4wZNwWqIr6lAJAgCMYBOMocMx7nciQUMztErVg4lRmnnoH4Yj0vE93UAQQjGAfgKMaM22BpswgcGjNOPQPxxZhxU/CQLSZxA0ALBYCjDHoK26BSbLG0GZCkyIwHBH8xTXUAIBgH4Cg/E7iFY2Ixe2ZXViZwA5IKE7iZgj8LmcQNAC0UAPXCw7jdKiy5Zc9ssMeqm7p54ticD0Dt0E3dFPxZSG0AIBgH4Cgy4zbI2FYv5plxXnxAXJEZNwW/7ZMZB0BLEICjmMDNTqBS4luKhMOYcSBJMWY8IPjdiOoAUKtg/NFHH1VeXp4yMzOVn5+vTz75pNr99+zZo7Fjx6pt27bKyMjQcccdpzfeeKNWBQbQsJhLmxEPVWGWb3uxXv6IHghAYiAzbvIygRuAIKnRHrBw4UIVFBRozpw5ys/P1+zZszV48GCtW7dOrVu3Dtu/vLxcZ599tlq3bq1Fixapffv2+vbbb9WiRYtYlB9AgjNHRxOMV2HMeAQOLW1GPQPxxZhxU/BnoUF9AK4XdTA+a9YsXXPNNRozZowkac6cOVqyZInmzp2rSZMmhe0/d+5c/fjjj/roo4+UlpYmScrLy6tbqQE0GIF1VJnALQhjme3FOntGPQMJIsYrJTRgwZ+FfmJxwPWi6rtXXl6uTz/9VIMGDao6gderQYMGaeXKlbbHvPbaa+rXr5/Gjh2rnJwcnXjiiZoxY4Z8Pl/E65SVlamkpMTyA6BhCiQnmcAtGN3UbcU8e0Y9Awkh1islNGCWzDj1AbheVC2U3bt3y+fzKScnx7I9JydHO3bssD3mm2++0aJFi+Tz+fTGG2/orrvu0syZM/XHP/4x4nUKCwvVvHlz8yc3NzeaYgJIIFXd1InGTXSfthfzpc2oZyAh0E3dZO2mDsDtHE8X+P1+tW7dWk8++aR69+6tESNG6I477tCcOXMiHjN58mQVFxebP1u3bnW6mAAc4vczgVsYJhaLIMZdWalnIDHEenLGBswygRu99gHXi2rMeKtWrZSSkqKdO3datu/cuVNt2rSxPaZt27ZKS0tTSkqKua179+7asWOHysvLlZ6eHnZMRkaGMjIyoikagARFbtIOS27ZYmkzIEkRjAdYljYjNw64XlTpgvT0dPXu3VtFRUXmNr/fr6KiIvXr18/2mDPOOEMbNmyQ31/19d/XX3+ttm3b2gbiAJJLYGkzLwFRFTK29jxkxoGkxNJmpuDPQiZwAxB1C6WgoEBPPfWU5s+fr6+++krXX3+9SktLzdnVR44cqcmTJ5v7X3/99frxxx81fvx4ff3111qyZIlmzJihsWPHxu5ZAEhcJCfDMZbZXszHjJsnjs35ANQOY8ZNTOAGIFjUS5uNGDFCu3bt0pQpU7Rjxw716tVLS5cuNSd127Jli7zeqhg/NzdXb775piZMmKCTTz5Z7du31/jx4zVx4sTYPQsACctvzqZOQGQiYxuBU5lxXntAXJEZN3nIjAMIEnUwLknjxo3TuHHjbB9bvnx52LZ+/frp448/rs2lADRwjImzQ3cBW2Z1MGYcSC6MGQ/m8VRUBZ+PAEjLAHCU34yHCIhMNEjtxTp7Zp6H1x4QV7GeD6KBMzvt81EAuB7BOABHGeYEbnEuSCIJtMDoph4i0GCP0emoZyAxxHylhIYtMGyLYBwALRQAjjKnKiMYD0L3aVssbQYkKbqpBzM7CvDlBOB6BOMAHGWwtFk4JnCzx9JmQHJiAjeLwLAtJnADQAsFgKPMnsLxLUZiYWkzezFf2ox6BhICS5tZVI0Zpz4AtyMYB+AoJnCzQcY2AjLjQFJiAjcLxowDCKCFAsBRgW/+icWDMZbZFmPGgSQV48kZGzgPQ+gBVCIYB+CoQGacMeNByNjaY8w4kJwYM27hNceME40DbkcLBYDDKjPjcS5FQmEscwQxTBdZzkE9A3HFmHELagNAAME4AEf56SkczszYUikWscyeBQfj1DMQZ4wZt6isDjLjAAjGATjKYAI3G4FK4S3YIqbZs+BgnHoG4irWKyU0cEzgBiCAFgoARxl0Uw9HC8xeLGc1oo6BxEE3dYuq76apD8DtCMYBOIoJ3GwwsVgEMezKGnwO6hmILyZws6iawC3OBQEQd7RQADiKpc2qQaVYxTRoZsw4kDhYyyuY2U+A6gBcj2AcgKMMMuPhyIzbi+XSZmTGgcRBZtzCw9JmACrRQgHgKHPMOLF4FZY2sxfLSZ5Y2gxIHIwZt4jl9BgAGjaCcQCO8pureBEQmciMR0BmHEhKZMYtvCxtBqASLRQAjiIHbIfF122ZQXOslzajnoGEQPApSfLwiQigEsE4AEf5mcAtnJkdolIsnBozTj0D8UW/bAsPmXEAlQjGATiLCdzCBRpgdJ8O4dA649QzEF8x7fXS8AU+D4nFAdBCAeAocwK3OJcjsdBN3Rbd1IEkRWbcDrUBgGAcgKP8ZtxJQGRiAjd7Me2mTmYcSBhM4GbhrawOuqkDoIUCwFEGSeBwLG1mj6XNgOTE0mYWgQnciMUBEIwDcFTgm38v8VAVMzNOpdiK+dJm1DMQV2TGLbxmr32iccDtCMYBOKoqB0xAVIXuArZi2p2cMeNA4mDMeLDAsC1qAwDBOABHBb759/JuU4XZ1O05sbQZdQzEH5lxC3NpMz/hOOB2tFIAOMqMO8mMV2HMuD1HxoxTx0DcMWbcgtoAEFCrYPzRRx9VXl6eMjMzlZ+fr08++aRGx/3973+Xx+PRsGHDanNZAA2QvyoaRwBjxiNwIjNOHQPxF8N7OwkEuqkzmzqAqIPxhQsXqqCgQFOnTtXq1avVs2dPDR48WD/88EO1x23evFm33nqrzjzzzFoXFkDDE2hreAmKgtBN3VZMs2fUMZAwPIwZD+YlNQ6gUtStlFmzZumaa67RmDFj1KNHD82ZM0eNGzfW3LlzIx7j8/l02WWXafr06erSpUudCgygYaGzsA26UNujmzqQnOimbmEubRbncgCIv6iC8fLycn366acaNGhQ1Qm8Xg0aNEgrV66MeNzdd9+t1q1b66qrrqrRdcrKylRSUmL5AdAwGSxtFo7JxSJgAjcgKcXyi7YkYE7gRn0ArhdVK2X37t3y+XzKycmxbM/JydGOHTtsj1mxYoX++te/6qmnnqrxdQoLC9W8eXPzJzc3N5piAkgg5pBxuqkHYWkzW2bgHMtu6tQxEH90Uw9mLm1GdQCu52jKYO/evbriiiv01FNPqVWrVjU+bvLkySouLjZ/tm7d6mApATgp8M0/MVEQsrb2WNoMSE4sbWbhJTMOoFJqNDu3atVKKSkp2rlzp2X7zp071aZNm7D9N27cqM2bN2vo0KHmNr+/4o04NTVV69atU9euXcOOy8jIUEZGRjRFA5CgqkbuEo2bGM9sjzHjQHJizLiF+b1jfIsBIAFElTJIT09X7969VVRUZG7z+/0qKipSv379wvbv1q2bvvjiC61Zs8b8Of/88zVw4ECtWbOG7ueAC/gZMx6OZbciiGVmnNnUgYRBZtzCa3ZTJxwH3C6qzLgkFRQUaNSoUerTp4/69u2r2bNnq7S0VGPGjJEkjRw5Uu3bt1dhYaEyMzN14oknWo5v0aKFJIVtB5CcDIbu2qBSbDmytFkMTgWgjhgzHsx8p6M6ANeLOhgfMWKEdu3apSlTpmjHjh3q1auXli5dak7qtmXLFnm9ZCIAWNFNPQhdqO3Fsu8mdQwkDrqpWzGBG4BKUQfjkjRu3DiNGzfO9rHly5dXe+y8efNqc0kADZTfX9lNne/oqtCFOgImcAOSEt3ULZjADUAArRQAjqpqapChrEI3dVssbQYkKbqpB6OfAIAAgnEAjmICNxtkbe2xtBmQnMiMWzCBG4AAWikAHMUEbjYYz2yPpc2A5GTehgSfUtD3jlQH4HoE4wAcZZiZcYIiE1nbCMiMA0mJzLiFp/Lz0E8wDrgerRQAjiI/aYfuArYYMw4kqRiulJAEqsaMUyGA2xGMA3BUYMy4h6CoCllbe4wZB5ITmXELL5lxAJVopQBwFGPGbTCe2R5jxoHkxDrjFlVjxqkPwO0IxgE4qiokIiiqwjcU9mLZYKeOgcTBjGXBeFsCEEAwDsBRLG1mw+xCTaVYxLIrK10ygMRBN3WLqm7qfDkBuB3BOABnEROFowu1vZgmxqljIGHQTd0WsTgAgnEAjvKztFk4M2vLW7AVE7gBSYnMuAUTuAEIoJUCwFEkKO3QXcAWS5sBSYox48GYwA1AAME4AEcFvvknMx6ErK09ljYDkhOZcYvA5yGxOABaKQAcZVRmKAnFg9BdwB5LmwHJiTHjFlW1QX0AbkcwDsBRBpnxcGRtIyAzDiSlWPZ6SQIexowDqEQrBYCjAmPiiMWDMZ7ZFmPGgSTFmPFgHqoDQCWCcQCOorOwDbpQ24tlC5U6BhIHX4pZ0E0dQADBOABH+c3MOI0xE12o7cVykifqGEgcTOBmwdJmAAJopQBwlEFvYRtUir1YTvJEHQOJg37Zwarms6M+ALcjGAfgKJY2s0HW1p4jS5vxugPijsy4BZlxAAG0BAE4jKXNwpANsWc22GNwLvMcvPKAuGNpMyuzowD1AbgdwTgAR/krEyFeL0GRicx49RgzDiQXMuMWZMYBBNBKAeAoZoutBl2orVjaDEhSjBkPRj8BAAEE4wAc5ScmCmdmh6gUCyfGjFPHQPzF8t5OAh66qQOoRDAOwFEGE7iFM6eY5y3YwuzKGsN1xqljIP5i2uul4Qt8HhKLA6CVAsBRBhO42aC7gD2WNgOSE93Ug1W901EfgNsRjANwFJlxG0wuZi+WkzxRx0DioJu6hYcJ3ABUqlUr5dFHH1VeXp4yMzOVn5+vTz75JOK+Tz31lM4880y1bNlSLVu21KBBg6rdH0ByCYyJIxYPYmaHqBQLTwyzZ9QxkDhY2swilm91ABq2qIPxhQsXqqCgQFOnTtXq1avVs2dPDR48WD/88IPt/suXL9cll1yiZcuWaeXKlcrNzdU555yj77//vs6FB5D4qiZwIygykbWNwIEJ3KhjIP5iOR9EEgis9OmnPgDXi7qVMmvWLF1zzTUaM2aMevTooTlz5qhx48aaO3eu7f4LFizQDTfcoF69eqlbt256+umn5ff7VVRUVOfCA0h85CftMJ7ZFkubAUmKVHAwD5+IACpFFYyXl5fr008/1aBBg6pO4PVq0KBBWrlyZY3OsX//fh06dEhHHXVUdCUF0CAFvvn30vaoQtbWnhNLm1HHQPzFcj6IJOCtrA4/g8YB10uNZufdu3fL5/MpJyfHsj0nJ0dr166t0TkmTpyodu3aWQL6UGVlZSorKzN/LykpiaaYABIJ3dTDMZ7ZXnDgbBh1y2qTgQMSB2PGQ1QubRbnUgCIv3pNGdx33336+9//rldeeUWZmZkR9yssLFTz5s3Nn9zc3HosJYBY8jOBWzgza0ulWAXVR12DaYNu6kDiYDb1YB7GjAOoFFUw3qpVK6WkpGjnzp2W7Tt37lSbNm2qPfbPf/6z7rvvPr311ls6+eSTq9138uTJKi4uNn+2bt0aTTEBJBCDzHhk1ImVpT7q2kgNvPDopg7EHRO4WXgZQg+gUlStlPT0dPXu3dsy+VpgMrZ+/fpFPO6BBx7QPffco6VLl6pPnz5HvE5GRoaysrIsPwAaJqMyKCLsDEI3dXseBzLj1DEQf3RTt/DQTR1ApajGjEtSQUGBRo0apT59+qhv376aPXu2SktLNWbMGEnSyJEj1b59exUWFkqS7r//fk2ZMkXPP/+88vLytGPHDklS06ZN1bRp0xg+FQCJKDA/jZcscBUmF7NnGTNex+6s1DGQOJjAzaIqM044Drhd1MH4iBEjtGvXLk2ZMkU7duxQr169tHTpUnNSty1btsjrrWr8PP744yovL9eFF15oOc/UqVM1bdq0upUeQMJj6K4dKsWeE93UqWMg/uiXHSwwbIvqABB1MC5J48aN07hx42wfW758ueX3zZs31+YSAJKEwdJm4cja2iMzDiSnWC5bmASYwA1AAK0UAI6qGrlLNG5iPLM9xowDyYkx4xaMGQcQQDAOwFEsbWaDrG0EwcE4mXEgaTCbuoWXzDiASrRSADiKpc3sMJ7ZliVwZsw4kDwYMx6MjgIAAgjGATjKzIzHuRwJxcz6UisWHgcy49QxEH+MGbcIfDlNZhwAwTiAeuHl3aaK2V2ASrGwTOAWozHjZMaB+DPvbYJPKei7CaoDcD1aggAcZcZEZCiDECjaY2kzIDkRfQZjAjcAAQTjABzFBG42zMnFqBQLRzLjfMwBcWdO4EY3dYkJ3ABUoZUCwFFM4GbDbH9RJxYsbQYkJ2Yss6CbOoAAgnEAjgp88+8lJqrCslv2nJjAjToG4o/MuIW38r3OIBoHXI9WCgBHVeUnicarMJ45slhl0KhjIHGQCg5GPwEAAQTjABxlkBkPR9Y2slgtgUQdA4mDzLgFS5sBCKCVAsBRrDBlg/HMkZmNdsaMA0mDMeMWjBkHEEAwDsBRfoKicMymXo1YZ8apYyBhEH1Kqhq25ac6ANcjGAfgqEBbg27qwVh2KyKzTmI1Zpw6BuIuVj1ekkTV5yH1AbgdrRQAjmJpMxv0FogsVv03qWMgcdBN3YJu6gACCMYBOIoJ3GwwuVhksZroiToGEgcTuFkwgRuAAFopABxl5icJxoMwq11ksV7arI6nARADpIKDkRkHEEAwDsBRgW/+6aYehKxtZGTGgeRDZtzC62ECNwAVaKUAcJRBgjKc2QCjVsLEbMy4ecK6nQdA3TFm3KKqNqgPwO0IxgE4KvDNv5fMeBWW3apGrIJxMuNAwiAzbhH4PKSbOgBaKQAcZZjd1ONckITCmPGIYpZBo46BxMEg6WBVHYCoD8DtCMYBOKqqmzpBkcnMDlEnYcxWaozGjFPHQPzF6r5OMowZB0AwDsBRgTFxLG0WxPyGgrfgMGZ31hitM04dA/Fn3odEn1JQN/U4lwNA/NFKAeAoM6YiGA9CF+rI6KYOJB+6qQejmzqAAIJxAI4KLG3GBG5BmFwsMpY2A5IPC2tbMIEbgABaKQAcRWLcBt0FIovZ0mbUMZAwWNrMwnyboz4A1yMYB+CoQEzkZdB4FbK2kZEZB5IPS5tZeCqjcT/VAbherVopjz76qPLy8pSZman8/Hx98skn1e7/4osvqlu3bsrMzNRJJ52kN954o1aFBdDwmEubxbkciYXxzJExZhxIPnRTD1b1Lkd9AG6XGu0BCxcuVEFBgebMmaP8/HzNnj1bgwcP1rp169S6deuw/T/66CNdcsklKiws1Hnnnafnn39ew4YN0+rVq3XiiSfG5Ekkgk27S/Xk+xv14Yb/6af95fJIapqRqkbpKTrsq3i7TUvx6EC5T3vLDstfg/UsUrwe8xzVHRe63/5DPjVOSznicbW9Xuj+LRqnK6tRqooPHFLx/kOOXS9wTMvG6TqpQ3NJ0uffFeun/eWOXa8u16zL9WpTr4l6vf3lPknSE+99o6ObZqhzqyZHLFPSI2sbWcwz4wTjQNyRGbf4sbRckvTe17t00rQ3zc9Ou8/V4PZkpM/aQDvljGNa6ZpfdOFzVuHt8urquK6PVae2bfR4lbM+2568Zit4jCincszPz9dpp52mRx55RJLk9/uVm5urG2+8UZMmTQrbf8SIESotLdXrr79ubvvZz36mXr16ac6cOTW6ZklJiZo3b67i4mJlZWVFU9x68d8Fk1Sydrn5/eZK/wnq5/2vTvBuVrHRxJLnae4pVYnRxPx3JIHjsoL2sTsudL9io4m2qZXaabe5XyyvF7q/IZnHBfYPbI/l9QLHfGdkK9ezyzyuxGiirUa2Onh2xfx6dblmXa5Xm3ptSNc7unG6mhilUqMWwYuQSweKpczmVf8OFtj3YND2SPtWd1yjFlLzDlLx1qrj7M7j5PXanixt/UTylUnt+0hpjaROZ0gDJ0c+rxssK5S+/VDa+k/JVy617SVlZkk/bg7/OxyVZ7898Pdp1ELyH5ZKvpey2ktHdaGOgXhZViitWyLt+EJKbSR16CPt+Ta699xIjx/pvSCSOF6v3OfXgUOHwz4z7T5X/+vPkySd4N1sbrM77jsj23K55t0G6ITL7otcniQX2i6XpA6eXZIUse1S28eiaXvWtI1eXRvLyXLWZ1uwIb1mnY5DowrGy8vL1bhxYy1atEjDhg0zt48aNUp79uzRq6++GnZMx44dVVBQoJtvvtncNnXqVC1evFifffZZja6byMH4pt2leuUvN6kgbVG8i2Iq9Weoibcs3sUAElNaY+nQ/sS63sA7pP631U95EtV7D0jL7nXu/NQxEB9O39sI8+rh09X8tIvUvFFavItS7/YcOKTif72gYakfxbsotmij25t1+EINv/kh5SVghtzpODSqbuq7d++Wz+dTTk6OZXtOTo7Wrl1re8yOHTts99+xY0fE65SVlamsrOqFWlJSEk0x69ULq7Zqjm+4UnVYN6UtjndxJImbHKhOfQbiNbkeQWKFQB040WinjoH46X+bVPydtHp+vEviGr9J/Uj6d2IGo/Ui6kG49Yc2eriZhy7UI77hOrRqqyYO6Rbv4tS7hHy5FhYWavr06fEuRo1899MBGZIe8/1G41Jflddj7WhgGNaVemozfLGmx4Xu5/T1Qvevr+tJkt+oOCC0vp26Xl2uWdd6qe3fI9GvB0kp6QSJwfrfJr13f0UX81jxplHHQLydO1P6998YM66afWba7VOT4/yGR/82jpEkpXo9ykhNqUNJG6aywz4drhyzfIpnQ8Q2W3X1WdvHqlPbNnq8yllfbcHDhlcP+4ZLqoip3CiqYLxVq1ZKSUnRzp07Ldt37typNm3a2B7Tpk2bqPaXpMmTJ6ugoMD8vaSkRLm5udEUtd50aNlIHknXpCyxveGDX5i1DVhqelzofk5fL3h/n+FRSi0C49pcT6pdEF6X69XlmnX5O9SmXhvK9aCKsdHvPUCwGPDeA7ENxCXJf4g6BuJtxV8IxCvV5DPTbp+aHOf1GFp+uKce8Q3XdQO6ujLLeP/StZqzfKPGpbys3mnrI+5XXX3W9rHq1LaNHo9y1mdbMNXj140pL+sR33B1aNmodidp4KKayjc9PV29e/dWUVGRuc3v96uoqEj9+vWzPaZfv36W/SXp7bffjri/JGVkZCgrK8vyk6gu6pOrG1Ne1i0JNGZ8j79xvV+ztoE4qlff9eqKv2Nm88S63rJ7K4JFt3NyXCl1DMQPY8br1S1pi3Rj6ssa0Scxk1hOS8R2ebB4tNGjVd9tQbe/ZqPupl5QUKBRo0apT58+6tu3r2bPnq3S0lKNGTNGkjRy5Ei1b99ehYWFkqTx48erf//+mjlzps4991z9/e9/16pVq/Tkk0/G9pnESedWTTS4R7ZWru2eWLOp+5lNva7Xq8s1G9Ls5vV9PWZTP1nq/Atp0/tVx/l9kc/pFn6flHemdZvHU/vZ1Ft0DD8/gPpnd28zm7qjs6kP7p6dkBNh1Qe7drmUYLOpH6GN7sbZ1N38mo16aTNJeuSRR/SnP/1JO3bsUK9evfTQQw8pPz9fkjRgwADl5eVp3rx55v4vvvii7rzzTm3evFnHHnusHnjgAf3617+u8fUSeTb1gM27S/Xk+99oxYbdlnXGG6en6pDPL0NSeopX+8sPa2/ZYRk1WIPPW7l2X+P01GqPC90vsIbhkY6r7fVC9w+sSVhy4JD2VK5J6MT1Ase0bJyukyvX/P6scs1vp65Xl2vW5Xq1qddEvZ43aC3Ja3/RxbVvtgAASOFtxsBnp93nanB7MtJnLZ+z4aKp47o+Vp3attHjVc76bHs2lNdsQi1tFi8NIRgHAAAAACQPp+PQqMaMAwAAAACAuiMYBwAAAACgniXkOuOhAj3pS0pK4lwSAAAAAIAbBOJPp0Z2N4hgfO/evZKUsGuNAwAAAACS0969e9W8eeyXx20QE7j5/X5t27ZNzZo1k6e2q8oDSGolJSXKzc3V1q1bmegRgGvxXggAsXsvNAxDe/fuVbt27eT1xn6Ed4PIjHu9XnXo0CHexQDQAGRlZdEABeB6vBcCQGzeC53IiAcwgRsAAAAAAPWMYBwAAAAAgHpGMA4gKWRkZGjq1KnKyMiId1EAIG54LwSAhvNe2CAmcAMAAAAAIJmQGQcAAAAAoJ4RjAMAAAAAUM8IxgEAAAAAqGcE4wAAAAAA1DOCcQAAAAAA6hnBOAAAAAAA9YxgHAAAAACAekYwDgAAAABAPSMYBwAAAACgnhGMAwAAAABQzwjGAQAAAACoZwTjAAAAAADUM4JxAAAAAADqGcE4ACDhzJs3Tx6PR6tWrYp3URLWgAEDNGDAgLiWwePxaNq0aXEtQ20lQv0BANyNYBwAAJd5/vnnNXv27Lhdf9u2bZo2bZrWrFnj6HW+/PJLTZs2TZs3b3b0OgAA1AbBOAAADdBbb72lt956q1bHJkIwPn369HoJxqdPn24bjNel/gAAiIXUeBcAAABELz09Pd5FaNCoPwBAvJEZBwDUu++//15XXXWV2rVrp4yMDHXu3FnXX3+9ysvLLfuVlZWpoKBA2dnZatKkiX77299q165dln1effVVnXvuuea5unbtqnvuuUc+n8+y34ABA3TiiSfqyy+/1MCBA9W4cWO1b99eDzzwQFj5vv32W51//vlq0qSJWrdurQkTJujNN9+Ux+PR8uXLLfv+85//1JAhQ9S8eXM1btxY/fv314cffmjZZ9q0afJ4PFq7dq0uuugiZWVl6eijj9b48eN18OBBy76HDx/WPffco65duyojI0N5eXm6/fbbVVZWFvZ8gsc8L1++XB6PRy+88ILuvfdedejQQZmZmfrlL3+pDRs2WI5bsmSJvv32W3k8Hnk8HuXl5dn+nYL/DhMmTFB2draaNWum888/X999953tvt9//72uvPJK5eTkKCMjQyeccILmzp1rKedpp50mSRozZoxZhnnz5kVVp4FrRXodzZs3T7/73e8kSQMHDjSvE/j72Y0Z/+GHH3TVVVcpJydHmZmZ6tmzp+bPn2/ZZ/PmzfJ4PPrzn/+sJ5980vw7nXbaafrXv/5VbT0CABCMzDgAoF5t27ZNffv21Z49e3TttdeqW7du+v7777Vo0SLt37/fkrG88cYb1bJlS02dOlWbN2/W7NmzNW7cOC1cuNDcZ968eWratKkKCgrUtGlTvfvuu5oyZYpKSkr0pz/9yXLtn376SUOGDNHw4cN10UUXadGiRZo4caJOOukk/epXv5IklZaW6qyzztL27ds1fvx4tWnTRs8//7yWLVsW9lzeffdd/epXv1Lv3r01depUeb1ePfPMMzrrrLP0wQcfqG/fvpb9L7roIuXl5amwsFAff/yxHnroIf3000969tlnzX2uvvpqzZ8/XxdeeKFuueUW/fOf/1RhYaG++uorvfLKK0es3/vuu09er1e33nqriouL9cADD+iyyy7TP//5T0nSHXfcoeLiYn333Xf6y1/+Iklq2rRptee8+uqr9dxzz+nSSy/V6aefrnfffVfnnntu2H47d+7Uz372M3k8Ho0bN07Z2dn6v//7P1111VUqKSnRzTffrO7du+vuu+/WlClTdO211+rMM8+UJJ1++ulR1emRXke/+MUvdNNNN+mhhx7S7bffru7du0uS+f9QBw4c0IABA7RhwwaNGzdOnTt31osvvqjRo0drz549Gj9+vGX/559/Xnv37tXvf/97eTwePfDAAxo+fLi++eYbpaWlHfHvBACADAAA6tHIkSMNr9dr/Otf/wp7zO/3G4ZhGM8884whyRg0aJC5zTAMY8KECUZKSoqxZ88ec9v+/fvDzvP73//eaNy4sXHw4EFzW//+/Q1JxrPPPmtuKysrM9q0aWNccMEF5raZM2cakozFixeb2w4cOGB069bNkGQsW7bMLOuxxx5rDB482FLG/fv3G507dzbOPvtsc9vUqVMNScb5559vKecNN9xgSDI+++wzwzAMY82aNYYk4+qrr7bsd+uttxqSjHfffdfyfPr372/+vmzZMkOS0b17d6OsrMzc/uCDDxqSjC+++MLcdu655xqdOnUKqzc7gTLdcMMNlu2XXnqpIcmYOnWque2qq64y2rZta+zevduy78UXX2w0b97c/Fv961//MiQZzzzzjGW/aOq0Jq+jF1980fI3CxZaf7NnzzYkGc8995y5rby83OjXr5/RtGlTo6SkxDAMw9i0aZMhyTj66KONH3/80dz31VdfNSQZ//jHP8KuBQCAHbqpAwDqjd/v1+LFizV06FD16dMn7HGPx2P5/dprr7VsO/PMM+Xz+fTtt9+a2xo1amT+e+/evdq9e7fOPPNM7d+/X2vXrrWcr2nTprr88svN39PT09W3b19988035ralS5eqffv2Ov/8881tmZmZuuaaayznWrNmjdavX69LL71U//vf/7R7927t3r1bpaWl+uUvf6n3339ffr/fcszYsWMtv994442SpDfeeMPy/4KCAst+t9xyiyRpyZIlOpIxY8ZYehcEMs/BzzEagTLddNNNlu0333yz5XfDMPTSSy9p6NChMgzDrI/du3dr8ODBKi4u1urVq6u9Vk3rNNrXUU2fZ5s2bXTJJZeY29LS0nTTTTdp3759eu+99yz7jxgxQi1btjR/r2s9AwDch27qAIB6s2vXLpWUlOjEE0+s0f4dO3a0/B4Ifn766Sdz23//+1/deeedevfdd1VSUmLZv7i42PJ7hw4dwgK1li1b6vPPPzd///bbb9W1a9ew/Y455hjL7+vXr5ckjRo1KmL5i4uLLQHbsccea3m8a9eu8nq95mzf3377rbxeb9i12rRpoxYtWli+hIikJnUWjUCZunbtatl+/PHHW37ftWuX9uzZoyeffFJPPvmk7bl++OGHaq9V0zotLy+P6nVUE99++62OPfZYeb3WPEWgW3to3ce6ngEA7kMwDgBIWCkpKbbbDcOQJO3Zs0f9+/dXVlaW7r77bnXt2lWZmZlavXq1Jk6cGJaZPtL5ohE495/+9Cf16tXLdp8jjcWOlMGtTWY3IJbPMRqB+rj88ssjBtMnn3xyjc5xpDr98ccfa1/QGIlXPQMAkgfBOACg3mRnZysrK0v/+c9/YnK+5cuX63//+59efvll/eIXvzC3b9q0qdbn7NSpk7788ksZhmEJioNnJJdkZoqzsrI0aNCgGp17/fr16ty5s+Wcfr/fnM28U6dO8vv9Wr9+vWWisZ07d2rPnj3q1KlTbZ+WRTTBfqBMGzdutGTD161bZ9kvMNO6z+c7Yn1Eun5N67Smr6Non+fnn38uv99vyY4HhjrEqu4BAAhgzDgAoN54vV4NGzZM//jHP7Rq1aqwx6PNKgayk8HHlZeX67HHHqt1GQcPHqzvv/9er732mrnt4MGDeuqppyz79e7dW127dtWf//xn7du3L+w8oUuwSdKjjz5q+f3hhx+WJHMm91//+teSpNmzZ1v2mzVrliTZzmBeG02aNAnrwh9JoGwPPfSQZXtoGVNSUnTBBRfopZdesg2Sg+ujSZMmkip6NgSraZ3W9HUU6Tp2fv3rX2vHjh2WmfoPHz6shx9+WE2bNlX//v2PeA4AAKJBZhwAUK9mzJiht956S/3799e1116r7t27a/v27XrxxRe1YsUKtWjRosbnOv3009WyZUuNGjVKN910kzwej/72t7/Vqavw73//ez3yyCO65JJLNH78eLVt21YLFixQZmampKpsq9fr1dNPP61f/epXOuGEEzRmzBi1b99e33//vZYtW6asrCz94x//sJx706ZNOv/88zVkyBCtXLnSXC6sZ8+ekqSePXtq1KhRevLJJ80u+J988onmz5+vYcOGaeDAgbV+XsF69+6thQsXqqCgQKeddpqaNm2qoUOH2u7bq1cvXXLJJXrsscdUXFys008/XUVFRWE9BaSKZdWWLVum/Px8XXPNNerRo4d+/PFHrV69Wu+8847Zvbxr165q0aKF5syZo2bNmqlJkybKz89X586da1ynNXkd9erVSykpKbr//vtVXFysjIwMnXXWWWrdunVY2a+99lo98cQTGj16tD799FPl5eVp0aJF+vDDDzV79mw1a9YsJnUPAIApTrO4AwBc7NtvvzVGjhxpZGdnGxkZGUaXLl2MsWPHmktyBZY2C122KrB8V/BSVR9++KHxs5/9zGjUqJHRrl0747bbbjPefPPNsP369+9vnHDCCWFlGTVqVNgyX998841x7rnnGo0aNTKys7ONW265xXjppZcMScbHH39s2fff//63MXz4cOPoo482MjIyjE6dOhkXXXSRUVRUZO4TWNrsyy+/NC688EKjWbNmRsuWLY1x48YZBw4csJzv0KFDxvTp043OnTsbaWlpRm5urjF58mTLMm2B52O3tNmLL75o2S+wFFfwMmL79u0zLr30UqNFixaGpCMuc3bgwAHjpptuMo4++mijSZMmxtChQ42tW7eGLW1mGIaxc+dOY+zYsUZubq6RlpZmtGnTxvjlL39pPPnkk5b9Xn31VaNHjx5GampqWPlqUqeGceTXkWEYxlNPPWV06dLFSElJsbwmQusvUPYxY8YYrVq1MtLT042TTjopbPm1QH3+6U9/Cqsnu/oAACASj2Ew0wgAAEcye/ZsTZgwQd99953at28f1bHTpk3T9OnTtWvXLrVq1cqhEgIAgIaEMeMAAIQ4cOCA5feDBw/qiSee0LHHHht1IA4AAGCHMeMAAIQYPny4OnbsqF69eqm4uFjPPfec1q5dqwULFsS7aAAAIEkQjAMAEGLw4MF6+umntWDBAvl8PvXo0UN///vfNWLEiHgXDQAAJAnGjAMAAAAAUM8YMw4AAAAAQD0jGAcAAAAAoJ41iDHjfr9f27ZtU7NmzeTxeOJdHAAAAABAkjMMQ3v37lW7du3k9cY+j90ggvFt27YpNzc33sUAAAAAALjM1q1b1aFDh5ift0EE482aNZNUUQlZWVlxLg0AAAAAINmVlJQoNzfXjEdjrUEE44Gu6VlZWQTjAAAAAIB649RQaSZwAwAAAACgnhGMAwAAAABQz6IOxt9//30NHTpU7dq1k8fj0eLFi494zPLly3XqqacqIyNDxxxzjObNm1eLogIAAAAAkByiHjNeWlqqnj176sorr9Tw4cOPuP+mTZt07rnn6rrrrtOCBQtUVFSkq6++Wm3bttXgwYNrVWgADcem3aV6YdVWfffTAXVo2UgX9clV51ZN4l2sekUdAADvhQHUQ3Lx+Xw6dOhQvIuBWkpLS1NKSkrcru8xDMOo9cEej1555RUNGzYs4j4TJ07UkiVL9J///MfcdvHFF2vPnj1aunRpja5TUlKi5s2bq7i4mAncgAbkhVVbNemlz+XxeGQYhvn/+y84Wb/r447lCkPrYHzqS/IZHuUOm1ZVB+89IPl90sDJ8S1sPC0rlLwpUv/bqrbVpV5ifT4AtVN5L77Q5BLzvXCs9yWleAw9ePgCV30eSHwuJhPDMLRjxw7t2bMn3kVBHbVo0UJt2rSxnaTN6TjU8dnUV65cqUGDBlm2DR48WDfffLPTlwYQR5t2l2rSS5/Lb0gKfOdX+f+JL32u0/KOUl6SZwLs6uCw36Nb0hbpwcV+bWk9VR03vSAtu1caeEd8Cxtv3pSKevAflk67Wvp0Xt3qJXC+g8XSOX+U3v8T9QzEQ+W9+OPhz5Rq/E6/9/5DBamLNPPQhfIb7vk8kPhcTDaBQLx169Zq3LixY7NtwzmGYWj//v364YcfJElt27at9zI4Hozv2LFDOTk5lm05OTkqKSnRgQMH1KhRo7BjysrKVFZWZv5eUlLidDEBxNgLq7ZWfDDZdL7xeDxauGqrJg7pFoeS1R+7OnjYVzG855a0RTLmvlyxceAd1gyuGwWe/7J7pffur/h3Xeql/23Szi+llY9IHz8uGT7qGYiH/rfpn2s36brtz+valCXyegzNPHSh+V7ols8Dic/FZOLz+cxA/Oijj453cVAHgVj0hx9+UOvWreu9y3pCzqZeWFio5s2bmz+5uXTbARqa7346oEijYAzD0Hc/HajnEtW/SHXwsG+4DEPySJI3jQAxILgeYlEvOT0q/m/4pJR06hmIk5WpP5MkeT2GyoxUMxCX3PN5IPG5mEwCY8QbN24c55IgFgJ/x3iM/Xc8GG/Tpo127txp2bZz505lZWXZZsUlafLkySouLjZ/tm7d6nQxAcRYh5aNInbZ8ng86tDS/v5PJpHq4MaUl2Vu9h+qGMsMaz3Eol42fVDxf49X8pVTz0CcDDz4tqSKhHCG57BuTHnZfMwtnwcSn4vJiK7pySGef0fHg/F+/fqpqKjIsu3tt99Wv379Ih6TkZGhrKwsyw+AhuWiPrnVZgBGuGCiGrs6uDHlZd2Stkh7jcpG16kjK7tmuzxQfO+BinrIqHy/P+WKutXLew9Im9+v+HfPSyq6qFPPQP177wH13PUPSdJ3RrZmHrpQt6QtMgNyt3weSHwuAggXdTC+b98+rVmzRmvWrJFUsXTZmjVrtGXLFkkVWe2RI0ea+1933XX65ptvdNttt2nt2rV67LHH9MILL2jChAmxeQYAElLnVk10/wUnyxv0ZaPXU/Fz/wUnu2KSmkAdBH/hmuo1NOvwhUrNbFaxoddlFYGi3xefQiYKf+WY7oxAvVxat3rx+6S8n1f82zAquqhTz0D98/sq3uckeTyGHvYN18xDFyrVY7jq80AK+kwI2pbi9biuHgBUiXoCt1WrVmngwIHm7wUFBZKkUaNGad68edq+fbsZmEtS586dtWTJEk2YMEEPPvigOnTooKeffpo1xgEX+F2fXJ2Wd5R++9iH+mn/IZ3To40m/aqbqxocv+uTq4w0r276f2uU6vXo4Bm3aUSfXDWad4pUJsnwM5ZZqlpubNXciv/XtV4GTpbe/aO0eUXFuSTqGYiHgZOlb5ZLaxYou2m6VFYxb8Y1Z3bWu/mdXPV5IFV8Jny/54Bmv7NerZqm63d9cjWiT67r6gHxcaTu2FOnTtW0adPqpzCQVItgfMCAARG72EjSvHnzbI/597//He2lACSBvFZNdHTTDP20/5CGndLelQ2O1s0yJUmN0lOqZsoNWdYGlWJZL+Y5qGMgrirvxdSgOKDg7OPVKL1+Zy1OFEc3SZckHZfTjNnToU27S/XCqq367qcD6tCykS7qk6vODrWVtm/fbv574cKFmjJlitatW2dua9q0qflvwzDk8/mUmur44luulpCzqQNILobLgyK/PxBgBm91d51EFst64QsPIDGE39eGi9/7zNpwbxWg0gurtuqXM5fryfe/0ZLPt+nJ97/RL2cu14urnJm8uk2bNuZP8+bN5fF4zN/Xrl2rZs2a6f/+7//Uu3dvZWRkaMWKFRo9erSGDRtmOc/NN9+sAQMGmL/7/X4VFhaqc+fOatSokXr27KlFixY58hySDV91AHBcoMHhd2nDw1dZAf7glleg63Tg/6gQy3qhjoHEYHMvuvXzQKr6gtZPNJ5UDMPQgUM1n5dk8/9KNemlzyvuhZBeYRNf+lwnts9Sp6NrliFvlJYSsxnBJ02apD//+c/q0qWLWrZsWaNjCgsL9dxzz2nOnDk69thj9f777+vyyy9Xdna2+vfvH5NyJSuCcQCOc3sWwFfZ8LImxsna2qKbOpB8bD4EqhvymOx4Z0pOBw751GPKmzE5l9+QfvXgihrv/+Xdg9U4PTZh3d13362zzz67xvuXlZVpxowZeuedd8zVsrp06aIVK1boiSeeIBg/AoJxAI7z22WGXcT2+ZO1tUdmHEg+ZMYtzJFLLv1MRGLr06dPVPtv2LBB+/fvDwvgy8vLdcopp8SyaEmJYByA48xkZ3yLETe+QDuUMeM1wJhxIPnY3Ncuvi0DQThvTcmlUVqKvry75qtFzXr7az2zYrM5lC1YisejMT/PU8HZx9X42rHSpIm1a7zX6w374ujQoUPmv/ft2ydJWrJkidq3b2/ZLyMjI2blSlYE4wAc5zcbHu5seZjd1IOfvvkNBVlbC8P2m4tanos6BhKCbWbcnZ8HUvA8Ku6tg2Tk8Xii6ip+WX4nzV2xyfYxQ4Yuz+8Us67ndZGdna3//Oc/lm1r1qxRWlqaJKlHjx7KyMjQli1b6JJeC8ymDsBxbh8ebX4ZYckKuby7QCSxnGCAMeNAYrD5EHDzXRn4LHBzHUDq3KqJ7r/gZHk9UorXY/n//RecnDBLwZ511llatWqVnn32Wa1fv15Tp061BOfNmjXTrbfeqgkTJmj+/PnauHGjVq9erYcffljz58+PY8kbhvh/3QIg6RmMGa/8f9BGxjPbc2TMuDtfd0DCMO/nqnvRrZ8HUtVngZvHzaPC7/rk6rS8o7QwaJ3xEX1yEyYQl6TBgwfrrrvu0m233aaDBw/qyiuv1MiRI/XFF1+Y+9xzzz3Kzs5WYWGhvvnmG7Vo0UKnnnqqbr/99jiWvGEgGAfgOGZTt+umT9bWHmPGgeRjkxl38W1puP1DERZ5rZpo4pBu9X7d0aNHa/To0ebvAwYMiDiccPr06Zo+fXrEc3k8Ho0fP17jx4+PdTGTHt3UATjO7T2yq7qpB3F73/1IWNoMSD4296Lh4vuSbuoAAgjGATjO7Uub2c6mTjd1eyxtBiSfynvQE3QvuvTjQBITuAGoQjAOwHFG2D/cxe8P7poZmiFyaaVERDd1IPnQTT0YS5sBCCAYB+A4t0/gFryGqBmXk7W1R2YcSD4sbWbBBG4AAgjGATjO7WPGfXaZccaM22PMOJB8bMeMu1fV25ybawGARDAOoB64fcy4n8x4zZEZB5KPzZdsfhenhf10UwdQiWAcgOPcvoqLJTMeNibapZUSEWPGgeTD+12wqtqgPgC3IxgH4Di/7Trb7hGcADIMWYNDsrZWht3U83U9F3UMxJXNbOpu7SklBc+jEueCAIg7gnEAjnN7TsQ6m7pCgnG31oqNWNeL26fxBxKF3ZhxF9+WjBkHEEAwDsB5Lu8tHDybekW3RIJxW9aF2GNxQpvzAqh/NkubxakkiSDQPd3NdYDkN3r0aA0bNsz8fcCAAbr55pvrvRzLly+Xx+PRnj176v3aNUEwDsBxbp/ALXjMuN9QSLdpd9aJvRh3349ll3cAtRfopq7g90L33pd+vidEHI0ePVoej0cej0fp6ek65phjdPfdd+vw4cOOXvfll1/WPffcU6N9Ez2AjqXUeBcAQPJz+wRu/tClzeimbi/m3dTdPkACSBBhvV48rn7ro5s64m3IkCF65plnVFZWpjfeeENjx45VWlqaJk+ebNmvvLxc6enpMbnmUUcdFZPzJBsy4wAc5/rMeOjSZsFZXyYXqxLremECNyAxBN2D3kAXbZd+HkhM4IYgywql9x6wbnvvgYrtDsrIyFCbNm3UqVMnXX/99Ro0aJBee+01s2v5vffeq3bt2un444+XJG3dulUXXXSRWrRooaOOOkq/+c1vtHnzZvN8Pp9PBQUFatGihY4++mjddtttYfd4aDf1srIyTZw4Ubm5ucrIyNAxxxyjv/71r9q8ebMGDhwoSWrZsqU8Ho9Gjx4tSfL7/SosLFTnzp3VqFEj9ezZU4sWLbJc54033tBxxx2nRo0aaeDAgZZyJiKCcQCOc3GbS1LIerqG+Z/gDZAU+3qhLyiQGKruQQ/jpVnaLFkZhlReGt2P4ZOW3Su9+8eK39/9Y8Xvhi/K89TttdSoUSOVl5dLkoqKirRu3Tq9/fbbev3113Xo0CENHjxYzZo10wcffKAPP/xQTZs21ZAhQ8xjZs6cqXnz5mnu3LlasWKFfvzxR73yyivVXnPkyJH6f//v/+mhhx7SV199pSeeeEJNmzZVbm6uXnrpJUnSunXrtH37dj344IOSpMLCQj377LOaM2eO/vvf/2rChAm6/PLL9d5770mq+NJg+PDhGjp0qNasWaOrr75akyZNqlPdOI1u6gAcF/iMIDNeWQdkxu2RGQeSU0hm3Cf3fh5IVV/Q+nlrSi6H9ksz2tXu2Pf/VPET6fcjuX2blN4k6ssahqGioiK9+eabuvHGG7Vr1y41adJETz/9tNk9/bnnnpPf79fTTz8tj8cjSXrmmWfUokULLV++XOecc45mz56tyZMna/jw4ZKkOXPm6M0334x43a+//lovvPCC3n77bQ0aNEiS1KVLF/PxQJf21q1bq0WLFpIqMukzZszQO++8o379+pnHrFixQk888YT69++vxx9/XF27dtXMmTMlSccff7y++OIL3X///VHXTX0hGAfgOHPmWJe2vXzBMabEmPFIGDMOJCfDJjPu4tvSxU8dCeL1119X06ZNdejQIfn9fl166aWaNm2axo4dq5NOOskyTvyzzz7Thg0b1KxZM8s5Dh48qI0bN6q4uFjbt29Xfn6++Vhqaqr69OkTcTjKmjVrlJKSov79+9e4zBs2bND+/ft19tlnW7aXl5frlFNOkSR99dVXlnJIMgP3REUwDsBx5mQ18S1G3AR/GFX8O8azhieLmM8yTzd1IDEQjAdjArcklda4IkMdrRV/qciCp6RLvnLpF3+Qfj4h+mtHYeDAgXr88ceVnp6udu3aKTW1KiRs0sSaYd+3b5969+6tBQsWhJ0nOzs7unJWatSoUdTH7Nu3T5K0ZMkStW/f3vJYRkZGrcqRCAjGATjO9RO4sbRZDTm1tBlfeABxZZMZd+vngRT8mRjngiC2PJ7ou4q/90BFID7wDqn/bZWTt91bEZj3v82Zcqoi4D7mmGNqtO+pp56qhQsXqnXr1srKyrLdp23btvrnP/+pX/ziF5Kkw4cP69NPP9Wpp55qu/9JJ50kv9+v9957z+ymHiyQmff5fOa2Hj16KCMjQ1u2bImYUe/evbtee+01y7aPP/74yE8yjpjADYDj3L60WfCYcUOhS5sRKJosY8bppg4kDZtgHEzgBkl+X1UgLlX8f+AdFdsTxGWXXaZWrVrpN7/5jT744ANt2rRJy5cv10033aTvvvtOkjR+/Hjdd999Wrx4sdauXasbbrih2jXC8/LyNGrUKF155ZVavHixec4XXnhBktSpUyd5PB69/vrr2rVrl/bt26dmzZrp1ltv1YQJEzR//nxt3LhRq1ev1sMPP6z58+dLkq677jqtX79ef/jDH7Ru3To9//zzmjdvntNVVCe1CsYfffRR5eXlKTMzU/n5+frkk0+q3X/27Nk6/vjj1ahRI+Xm5mrChAk6ePBgrQoMoOFxe5c86zrjin3QmSxi/SUFmXEgMdgsbUZmnMw4JA2cHJ4B739bxfYE0bhxY73//vvq2LGjhg8fru7du+uqq67SwYMHzUz5LbfcoiuuuEKjRo1Sv3791KxZM/32t7+t9ryPP/64LrzwQt1www3q1q2brrnmGpWWlkqS2rdvr+nTp2vSpEnKycnRuHHjJEn33HOP7rrrLhUWFqp79+4aMmSIlixZos6dO0uSOnbsqJdeekmLFy9Wz549NWfOHM2YMcPB2qm7qLupL1y4UAUFBZozZ47y8/M1e/ZsDR48WOvWrVPr1q3D9n/++ec1adIkzZ07V6effrq+/vprjR49Wh6PR7NmzYrJkwCQuKzjpeNYkDjyVVsHLq0UWw7VhVtfeEDCYMx4sKovqONbDrhTdZniSI+1adPGzD7bSU1N1ezZszV79uyI+yxfvtzye2ZmpmbNmhUxHrzrrrt01113WbZ5PB6NHz9e48ePj3id8847T+edd55l25gxYyLuH29RZ8ZnzZqla665RmPGjFGPHj00Z84cNW7cWHPnzrXd/6OPPtIZZ5yhSy+9VHl5eTrnnHN0ySWXHDGbDiA5BH/z79YsQPBs6uFLm7m0Uuw4lhmnjoG4IjNuEXjubu0tBqBKVMF4eXm5Pv30U8tAe6/Xq0GDBmnlypW2x5x++un69NNPzeD7m2++0RtvvKFf//rXEa9TVlamkpISyw+AhskIHS/tQpZu6hJLm0XC0mZAcrLcz0bQf92JdyYAAVF1U9+9e7d8Pp9ycnIs23NycrR27VrbYy699FLt3r1bP//5z2UYhg4fPqzrrrtOt99+e8TrFBYWavr06dEUDUCCIjNu7abu94dmxhnPbIp1vTBmHEgMNplxN2eFDYPeAQAqOD6b+vLlyzVjxgw99thjWr16tV5++WUtWbJE99xzT8RjJk+erOLiYvNn69atThcTgEMs2XCXNjzCG1zhWSJIsa8XBmYCiYEx48EYMw4gIKrMeKtWrZSSkqKdO3datu/cuVNt2rSxPeauu+7SFVdcoauvvlpSxbpypaWluvbaa3XHHXfI6w3/PiAjI6NBL94OoIql53H8ihFX4bOps7SZLZY2A5KTZWmzyk3xKUlCcPsKIwCqRJUZT09PV+/evVVUVGRu8/v9KioqUr9+/WyP2b9/f1jAnZKSIok3IcANgm9zt3bJ84XWARO42WNpMyA52U3g5tZxSwqewC3OBUGd+f18viSDeP4do17arKCgQKNGjVKfPn3Ut29fzZ49W6WlpeaU8SNHjlT79u1VWFgoSRo6dKhmzZqlU045Rfn5+dqwYYPuuusuDR061AzKASSv4G7qbm14hE3gJjLjtix1QTd1IHnYdFOPV1ESAH12Gr709HR5vV5t27ZN2dnZSk9Pl8fjOfKBSCiGYai8vFy7du2S1+tVenp6vZch6mB8xIgR2rVrl6ZMmaIdO3aoV69eWrp0qTmp25YtWyyZ8DvvvFMej0d33nmnvv/+e2VnZ2vo0KG69957Y/csACQsJnCTfEFPPCwzTnMsCJlxICkF3YOBYNytPaWkqufu5jpo6Lxerzp37qzt27dr27Zt8S4O6qhx48bq2LGj7fBpp0UdjEvSuHHjNG7cONvHQhd0T01N1dSpUzV16tTaXApAA8fSZtbZ1MPHjLuzTmyxtBmQnIJuQY/ZYyU+RUkIdNpJCunp6erYsaMOHz4sn88X7+KgllJSUpSamhq3ng21CsYBoKb8xJ0hE7ixtFlELG0GJCe7MeMu/TyQyIwnE4/Ho7S0NKWlpcW7KGig6j8XD8BdLMG4Oxselsx40H8V9m+3c6heXPq6AxKH3Zhx996X9NkBEEAwDsBRwd/8uzUTEj5mnAncbMV8abNAZtylLzwgUQRnxj1kxgPP3a1fUAOoQjAOwFGWXKdL2x1hQ6FZfN0eY8aB5GSE93pxcyAaeO4urgIAlQjGATiKCdysmfGK6iAzbsumwV7HE9qcF0D9Y2mzYHxNCCCAYByAo5jAzTpmnKXNqsPSZkBSspnAjcw4E7gBIBgH4LDgbLhbG1/+0MGRjBm3F/Mx4+SfgIRg2GTGXXxbGnTaAVCJYByAo4IbG26dsKfazDitsSqx/pKCzDiQGFjazCI4I+7WL6kBVCAYB+Ao61xl7mx0+BkzXjMx775P+glIDHaZcffel7GeqxJAw0UwDsBRLG3GmPGaIzMOJKWge9BDZtzy3Bk3DrgbwTgAR7G0meQL7pUukRaJhKXNgORkGTNuboxHSRKEYfMvAG5EMA7AUdYu2u5sdoTVgWXMOFlbU6zrhcw4kBiC3vu9qrgfyYwH/u3iigBAMA6g/ri1zWGdrEcK6S9Q38VJYKwzDiSn8My4m29LI+wzAYBbEYwDcBQTuFnHjId3Uydra4r50mZh/wAQD3ZLm7n4vnTvMwcQimAcgKOYwM3aTd3vNxgzHknMx4wHuqlTx0BcsbSZBd3UAQQQjANwFBO42WTGRTBuy1IXdFMHkkf4ve3WOUQkuqkDqEIwDsBR1vHS7mx1+IN6X7O0WXVY2gxISjaZcZd+HEiyPncy44C7EYwDcJR1zLg7+fwhlcCYcXsxHzPO0mZAQmDMuEXwc3dvLQCQCMYBOMywjBl3Z7PDFzpuPtZBZ7KI9ZcUZMaBxGA3ZtzFt6Wf1S0BVCIYB+AoxoyHrDNekRoPetSllWKLMeNAcrLLjLuXNTPu5poAQDAOwFF+MuNHyIyTFjHFul7IjAOJIegeDATjbv08kEJnU49fOQDEH8E4AEcxZjwkM26wtFlEMV/ajDHjQEKwGTPu6tvS8lbn5ooAQDAOwFExX62qAQqdvy3ms4Yni5jPMk83dSAxMIFbMCZwAxBAMA7AUXRTt86mbrC0WTWYwA1ISkHv/eYEbi5+67N2U3dxRQAgGAdQf9za5ggeM26wtFlkMV/aLOwfAOLB0k09bJPrGIzfAlCJYByAo8iMW8eMs7RZNWI+Zpx6BhJC8NJmnop/u/XzQGICNwBVCMYBOIoEQGhmPGRpMxc3SMM4NWZcop6BuAq//9x8R1qnUnFzTQCoVTD+6KOPKi8vT5mZmcrPz9cnn3xS7f579uzR2LFj1bZtW2VkZOi4447TG2+8UasCA2hY/GGBqLsYhmGJA/2h3dRpiAVxaMx4rM4HoHaCM+Oq+LcbPw8CDCOktxQA10qN9oCFCxeqoKBAc+bMUX5+vmbPnq3Bgwdr3bp1at26ddj+5eXlOvvss9W6dWstWrRI7du317fffqsWLVrEovwAEpwl7HRho8MX1tJiabOIYv1i4UsPIDEwZtzC+hHg4ooAEH0wPmvWLF1zzTUaM2aMJGnOnDlasmSJ5s6dq0mTJoXtP3fuXP3444/66KOPlJaWJknKy8urW6kBNBiGy8eMh8bi4WPGydiaYl0v1DOQGGwy4278PAiw9hiLY0EAxF1U3dTLy8v16aefatCgQVUn8Ho1aNAgrVy50vaY1157Tf369dPYsWOVk5OjE088UTNmzJDP54t4nbKyMpWUlFh+ADRMbh8zHtrgrPiVjK29WNcLPRCAxEBmPBidowAERBWM7969Wz6fTzk5OZbtOTk52rFjh+0x33zzjRYtWiSfz6c33nhDd911l2bOnKk//vGPEa9TWFio5s2bmz+5ubnRFBNAAqGbekgwHtZNnYytKeZLm/GlB5AQLN3UK/7t5juSCdwABDg+m7rf71fr1q315JNPqnfv3hoxYoTuuOMOzZkzJ+IxkydPVnFxsfmzdetWp4sJwCHWZb3c1+jwhTxnljarRqy/pKCbOpAYgu4/Mxh38XsfE7gBCIhqzHirVq2UkpKinTt3Wrbv3LlTbdq0sT2mbdu2SktLU0pKirmte/fu2rFjh8rLy5Wenh52TEZGhjIyMqIpGoAE5fZ2hj80Mx62tBlBoomlzYAkZZMZd/EtyQRuAAKiyoynp6erd+/eKioqMrf5/X4VFRWpX79+tsecccYZ2rBhg/z+qkbW119/rbZt29oG4gCSi9/lE7iFdVMPzYy7/uuKYGTGgaQU9N7vrbzP3fh5EOAnMw6gUtTd1AsKCvTUU09p/vz5+uqrr3T99dertLTUnF195MiRmjx5srn/9ddfrx9//FHjx4/X119/rSVLlmjGjBkaO3Zs7J4FgMTl8uRkaDd1xoxXI+ZjxiP+AqA+MWbcIvRTAYB7Rb202YgRI7Rr1y5NmTJFO3bsUK9evbR06VJzUrctW7bI662K8XNzc/Xmm29qwoQJOvnkk9W+fXuNHz9eEydOjN2zAJCwgr/1d2MmxO+3+Z0x4/YYMw4kJ8vSZmTGyYwDCIg6GJekcePGady4cbaPLV++PGxbv3799PHHH9fmUgAauOCZYt3Y9grPjFf9t+KfBImmmNeFy7tlAAkjfGUDV9+SvDUBqOT4bOoA3M3v8kZH6ARufsNwZ0XUiJOZceociBubzLibJy5z+1wqAKoQjANwVHCDy43rqYY1tAyFTqVbr+VJaLGuF9YZBxKD3ZhxF9+SlncmF9cDAIJxAA5ze6MjbDZ1sbRZRDEPnvnSA0gMwcF46Bb3sXzv6OqaAEAwDsBRhsu744U+Zz9Lm1WDCdyApGTppl7xbzd+HgQEP3cXVwMAEYwDcJg1A+A+vpAY0Ajrpk6QaIr50mZ0UwcSgmGTGXfxLclIJQABBOMAHGVd2ix+5YiX0G7qFRO4MbGYLZY2A5JT0P3nqcyMu3kCN7f3GANQhWAcgKMMl6cAQhtaLG1WjZh333f3aw9IHIwZD0afHQABBOMAHEVmPCQYD82M0xQLQmYcSEpBX4YxZpylzQBUIRgH4DB3L23mC82MM2Y8slh23w873n2vPSBhMGbcwuUdxgAEIRgH4ChLZtyFcWfouEjGjFcjll9ShK3v7sIXH5AogmdT9wQy4/EqTPz5LcG4iysCAME4AGcxm7r194r6YMSgvVjWRWgwTj0D8RN+/7mxp1QVw+ZfANyIYByAo4IbXG7MAISNGZfooxgJ3dSB5GTppl75bxffknwEAAggGAfgKL/LGx1hs6nTTT2ymHZTt+2SACAegrupV0bhbp64jAncAAQQjANwVHA23I3dEsNnUw/ZgbHMVWI6yzzd1IHEEZ4Zd/MtaRmo5OJ6AEAwDsBhwQ0NN07YEzqbetgEbi78gqJGYp4Z50sPIG5sM+PxKkz8+YOevBuHbwGoQjAOwFFuHzPuP+KYcYJEE2PGgeRkM2bcjT2lApjCE0AAwTgAR/ldPjw6tJs6S5tVw9Ex43zpAcRN0P1HN/XQHmMurggABOMAnOX2DED4BG6SpSYIEqswZhxIUnZjxt17T1rmUnFvNQAQwTgAh7l91tjQcZFhs6m78iuKSMiMA0mJMeMWfjLjACoRjANwlsuXNrOdTZ1FZu3Fsl4YMw4kjqDbjzHjIXOpxLEcAOKPYByAo6yNDvc1O8K6qQf9t+KfZGxNdFMHklRwN/XKLS6+JS3P3cX1AIBgHIDDLN3xXBh3HnECN1piQWLZTT00GHfhiw9IFJYJ3Cr+7eZu6kzgBiCAYByAo9zezjhyN3WCRJNllvm6notu6kDCMMIz426+J61LfsaxIADijmAcgKOYwC00GA9d2qyeC5TIWNoMSE6WCdwqM+MuviWZwA1AAME4AEdZhsa5sM3hC40Jg/5b8U8Xt0hDOTpmvI6nA1AHNmPGXXxTWpY2i2M5AMQfwTgARxkuz4z7Qp5zxZhxZu+xR2YcSEosbWYR/NzdvN46AIJxAA6z9DyOXzHixs+Y8ZpjaTMgOVnGjBuhm1wlNPh2az0AqFCrYPzRRx9VXl6eMjMzlZ+fr08++aRGx/3973+Xx+PRsGHDanNZAA1QcDbcjRmA0N4AfkMhY8bdVycRMWYcSE6W2dQDwbg73/tCn7abewgAqEUwvnDhQhUUFGjq1KlavXq1evbsqcGDB+uHH36o9rjNmzfr1ltv1ZlnnlnrwgJoeGKZ7GyIwmZTlyHGjEfAOuNAkrLJjMerKHEW3mfHrTUBQKpFMD5r1ixdc801GjNmjHr06KE5c+aocePGmjt3bsRjfD6fLrvsMk2fPl1dunSpU4EBNCxuHx0dNk4+tJu6K2slErqpA0nJtpu6O+9JuqkDCBZVMF5eXq5PP/1UgwYNqjqB16tBgwZp5cqVEY+7++671bp1a1111VU1uk5ZWZlKSkosPwAaJtdP4BaS+PaHLW3mvjqJiG7qQHIKurfdPoFb6PN24+cigCpRBeO7d++Wz+dTTk6OZXtOTo527Nhhe8yKFSv017/+VU899VSNr1NYWKjmzZubP7m5udEUE0ACcXs39fB1xiW6qUdAN3UgSQVlxj1u76bu1mcOwI6js6nv3btXV1xxhZ566im1atWqxsdNnjxZxcXF5s/WrVsdLCUAJ/ldnxk/wgRuNMyCkBkHkpLNBG5u/DyQ7CZwc2c9AKiQGs3OrVq1UkpKinbu3GnZvnPnTrVp0yZs/40bN2rz5s0aOnSouc3vr3hDTk1N1bp169S1a9ew4zIyMpSRkRFN0QAkKCPiL+5gO4EbS5vZs3Tfr+u5GDMOJAybMeNuvSXDphFxaT0AqBBVZjw9PV29e/dWUVGRuc3v96uoqEj9+vUL279bt2764osvtGbNGvPn/PPP18CBA7VmzRq6nwMu4PbMuG039VgGnckkpmPGQ1u8fOkBxE3Q/ed1eWbcdrlLAK4VVWZckgoKCjRq1Cj16dNHffv21ezZs1VaWqoxY8ZIkkaOHKn27dursLBQmZmZOvHEEy3Ht2jRQpLCtgNITpb4Kn7FiJvwYJylzSJizDiQpOxmU49XWeIrrM+OWysCgKRaBOMjRozQrl27NGXKFO3YsUO9evXS0qVLzUndtmzZIq/X0aHoABooN7Y5QmdTr5i/jaXN7LG0GZCULN3UKze59J4MW9osTuUAkBiiDsYlady4cRo3bpztY8uXL6/22Hnz5tXmkgAaKH9QHzw3dksM75LImPGIHF3azH2vPSBh2HZTj1dh4iv0eZMZB9yNFDYAR7k9Bxw2gVvY0mZurJUIYtpjgG7qQOIIv7dde0vy1gQgCME4AEcFZ4bdmAFgabNosLQZkJRsMuNu/DyQmMANgBXBOABHGS5PAgcaXt7AQEmWNovMMss8Y8aBpGGztJlb78jwdya31gQAiWAcgMOCsx9uHDMeyIynVEbjfr9iG3QmE0fHjPOlBxA3Qfc2S5uRGQdQhWAcgKPcPmY80PAKBOMVWRAy47ZY2gxIUixtFhDeacelFQFAEsE4AIdZx4zHsSBxEsiMp1Uu+ciY8eqQGQeSUtD953F5Zjx0rDyZccDdCMYBOCq0veW2SXsCDa2UlMrMuKHYdsdOJpbu+3U9F2PGgYRhM2bcrcIT4+6uD8DtCMYBOMrtPfIC66ynVmbGw7upx6FQiYqlzYAkRTf1gLAvqONTDAAJgmAcgKPCJ6txV9PDZwSC8eDMeHAGmMy4KZb1EtbipZ6BuLFZ2sxtnwUBTOAGIBjBOABnuTwLEBgznmp2UzdinAFOJjFcB49u6kDisFvazKW3JN3UAQQjGAfgKLdnxv0hmXF/2Jhxd9VHtRxd2ox6BuKGzLjJH5IKd2k1AKhEMA7AUeETuMWnHPFSlRkPjBmv+m/FP+k+bWLMOJCkwu9t7sgKBjUBuBrBOABHhY6Hc1tM5KuMtasy4wZLm0XE0mZAUrLJjLu1ezZjxgEEIxgH4KjQb/3dlgUINLxSKoNxsbRZZJYJ3BgzDiQNy5jxsE2u4vbeYgCsCMYBOCq0oeG2LIDZTT1SZpyWWBVHx4zzpQcQP1X3tlcV96Jrx4y7fB4VAFYE4wAcFdoV0W1dE0Mz4xW/khm3FdPu+6SfgIQQcu95gjoJuZFbnzcAewTjABzl9s7C5mzq5gRuLG0WGUubAUknNBh3+9JmYd8TurQiAEgiGAfgsNAueG5LBId3U1dIN3WXVUh1YpkZp5s6kBhC7j2PyydwC33ebhu6BcCKYByAo8KyAC7LUPoDs6kHMuOh3dSrNiKmMxtRx0BiiJAZj0dREkBYnx23VgQASQTjABwW+q2/27IAPsOaGTfCljYTrbGAWGazyYwDiSHk3gssbebWicuYwA1AMIJxAA5z9wRuod3UKxLjjGe2F8Nsdtih1DEQF4wZtwjvLQbAzQjGATjKH5KQdFtmvGoCtwhLm0nubZWGCmulkhkHGryImfF4FCb+wuZR4f0fcDWCcQCOCh0j7rYx44HMeIo3eMx4CALFCmH1wJhxoOGz7wnk1iA0plNjAGjwCMYBOCo0++G2hkf4bOo2mXGXfUFRYzHNjFPHQFxEyIy79ZYMfd6MGQfcjWAcgKPcngUIPN+UymDcstH8ncy4pNgG0IzLBxJDpDHjLr0nw3uLAXAzgnEAjnJ9N/XKhmhaSmA2dYku1BHENICmjoHEEBqMV2516S3p9i+oAVgRjANwVHiXvPiUI1785pjx6iZwIzMuiaXNgGQU1k294ne3ds9mAjcAwWoVjD/66KPKy8tTZmam8vPz9cknn0Tc96mnntKZZ56pli1bqmXLlho0aFC1+wNILqENDbc1PKrWGQ+awI0u1BHEcmkz6hhICGHd1Cs3139JEgLvTACCRR2ML1y4UAUFBZo6dapWr16tnj17avDgwfrhhx9s91++fLkuueQSLVu2TCtXrlRubq7OOeccff/993UuPIDExwRuNZjAjaxtBTLjQPIJC8b9dptdI/QLab/buosBsIg6GJ81a5auueYajRkzRj169NCcOXPUuHFjzZ0713b/BQsW6IYbblCvXr3UrVs3Pf300/L7/SoqKqpz4QEkvrAsgMvaHYGGVmpKZWY86L8mt1VKJIwZB5JQpDHj7rwnw8aMx6cYABJEVMF4eXm5Pv30Uw0aNKjqBF6vBg0apJUrV9boHPv379ehQ4d01FFHRVdSAA1S6Pg4t40TrOqmHpjAjaXNIotlN3Uy40BCiDhmPB6Fib/Q5+22z0QAVqnR7Lx79275fD7l5ORYtufk5Gjt2rU1OsfEiRPVrl07S0AfqqysTGVlZebvJSUl0RQTQCJxeRbAV9kOTfEGzabOdLr2WNoMSD4Rx4y7854Mn0clTgUBkBDqdTb1++67T3//+9/1yiuvKDMzM+J+hYWFat68ufmTm5tbj6UEEEtuz4wbIZlx+zHj7qqTiGK5/jpruQOJIVJm3KW3ZPg8Krz/A24WVTDeqlUrpaSkaOfOnZbtO3fuVJs2bao99s9//rPuu+8+vfXWWzr55JOr3Xfy5MkqLi42f7Zu3RpNMQEkELcngQPd1FMC64zb7USgWCGm3fdd/sIDEgb3XrDQHgHUDuBuUQXj6enp6t27t2XytcBkbP369Yt43AMPPKB77rlHS5cuVZ8+fY54nYyMDGVlZVl+ADRM4V0R3dX0CMymnsbSZtGjmzrQ8IV1U6+cxtKtX5DxPSGAIFGNGZekgoICjRo1Sn369FHfvn01e/ZslZaWasyYMZKkkSNHqn379iosLJQk3X///ZoyZYqef/555eXlaceOHZKkpk2bqmnTpjF8KgASUfhkNfEpR7wEZlNPqW4CNzLjFRxd2sxlLzwgUYR1U6+4F932WRDABG4AgkUdjI8YMUK7du3SlClTtGPHDvXq1UtLly41J3XbsmWLvN6qhPvjjz+u8vJyXXjhhZbzTJ06VdOmTatb6QEkPLqpV2bGLd3UXV4pkbC0GZCEImTGXdpbhW7qAIJFHYxL0rhx4zRu3Djbx5YvX275ffPmzbW5BIAkEdoV0W1ZAL85m3rFl5T2E7iRGZfkcGacOgbiIjQz7iEzHsy13fUBSKrn2dQBuE9YrtNl7Q5znfGUapY2IzdSKZbrjFPHQEKIOGY8HoWJP5Y2AxCMYByAo9y+tFlgAreqpc3EeOZIYrq0GZlxICGE3Htun8At9Gm77TMRgBXBOABHubmd4Q/qj2iZwC0sA0ygKImlzQAX8NjPnOEaYWPG3VoRACQRjANwmJsz48HPNdWytFksg85kQmYcSDoRZ1N35/ueP+StyK1j5wFUIBgHUK/c1P7yGTaZcRmx7Y6dTGLZfZ8x40BiYMy4Rfg7k0srAoAkgnEADguLO+NTjLgIzoCkBU/gRhdqezGtB+oYSAyhwXjlVpfek2HP253VAKASwTgAR7m5m7pdZtx+aTP31Em1Yrq0Gb0PgIQQNoFbxe9ufdsL7Zbups9EAOEIxgE4KiwmclG7wxfU6kpLqRwzbv4nmIsqpVosbQYknbBu6pWb678kCcKo5jcAbkMwDsBRod/6u6lrov1s6mJysUgcXdrMPa87IKGETeBW8btbM8LhmfH4lANAYiAYB+AoN+cng7upp1a7tJmbaqUaMc1mU8dAYog0Zrz+S5IIwnuLubQiAEgiGAfgsNCGht9FaYDAc/V6JI8nMGZcZMYjiemYceoYSAgsbWYR3lssTgUBkBAIxgE4ys0TxwYy4ylejypjcfulzVxVK9VhzDiQdCItbRaPsiQAljYDEIxgHICj3DybeqATgMfjMbtm+v1iPHMkjo4ZJzMOxEXYbOqBdcbd+b4X3lssTgUBkBAIxgE4ys0Thwe6qad4PPIGUuOSwjPAtMYk2dQDY8aBhi9CZtylt2R4bzGXVgQASQTjABzm6m7qgWA8uJu6QTf1yOimDiSdkFuPbuqMGQdQhWAcgKPCuuS5qOURGDPu9cjMjDOBWzWYwA1IPkzgZhHaLd1Fc5oCsEEwDsBRYflJFzU8/EGZ8YCKrAhdqG2xtBmQhOimHiz8abu0IgBIIhgH4DA3T+AWPJt69Zlx99RJtWJZL2TGgcQQYQI3yZ2TuIV/JsapIAASAsE4AEcxZryii3rVmHG5e1a7asVyzPgRNwCoDxGWNrN5yB3C3ubcWAkAAgjGATgq9Ft/NzU8AmMDgzPjFRO4kbW1xdJmQPKJMGZccldPqQAy4wCCEYwDcFRo8O2mtlfVBG5BmfGg/5rcVCnVcXTMeB1OBaAOqsmM13dREgDrPAAIRjAOwFGh8ZWbsgCBDIjXK3mCt8V0Pe1kQmYcSDqhmXEPmfFgbuotBiAcwTgAR4Wvqeqehoc5m7rHI4/ZTV2x7Y6dTGI6gRv5JyAhVHMfu+jjwBT29u/COgBQhWAcgKOYwE3yeoMncGNps4hY2gxIQpG7qbtR+Lucu+sDcDuCcQCOcnOXPHNpM0/wBG6iC3UksawX6hhIDEzgZhH6GejnrQlwNYJxAI4KywK4qO0VPJt6YMy4IdGFOqJYLm1GHQMJgaXNLMJ7i7mwEgCYCMYBOCrQ8PBWRqNumsAteDb1QGbcMoGbp/ItmKxthVjWC3UMJIbK90Gj8l50e2bcnNjThZ+JAMLVKhh/9NFHlZeXp8zMTOXn5+uTTz6pdv8XX3xR3bp1U2Zmpk466SS98cYbtSosgIYn0CUvpbLl4aYsgDmBm2XMuGRmaT0pwRthhNRLLMaMU8dAnFnvRdcvbVb5pAOfia6sBACm1GgPWLhwoQoKCjRnzhzl5+dr9uzZGjx4sNatW6fWrVuH7f/RRx/pkksuUWFhoc477zw9//zzGjZsmFavXq0TTzwxJk8iEWzaXaon39+oDzf8Tz/tL5dHUtOMVDVKT9FhX0X4kZbi0YFyn/aWHTYb6dVJ8XrMc1R3XOh++w/51Dgt5YjH1fZ6ofu3aJyurEapKj5wSMX7Dzl2vcAxLRun66QOzSVJn39XrJ/2lzt2vbpcsy7Xq029Jur19pf7JEmHfBWP3bboc81Y8lXU90Zdn1+090csrvdjabkkad2OvZr62n8kSQcO+VRyoFxZkg76pEyPdM38T/RPb3mDe36B44Lf7yKdqybX+8hTpuaeqnq5dv4n+th7yPa4I53vD95vNFpSmd+jDElv/nebCr5YWqvnF3zeSI/Hsj7ddL1I+zr9uenk9WJ1P9T0udX2b1qXuozmsV5lq/SYpLLK+zo4GD+98F3LF7aJ9Np06nplh62fif/e+qNOmvZmxHPX9PXUsnG6zjimla75RRd1btXkiOVJdqHt8tq8dmv6WHWc+Ix2spz12fbkNVvBY0Q5m1J+fr5OO+00PfLII5Ikv9+v3Nxc3XjjjZo0aVLY/iNGjFBpaalef/11c9vPfvYz9erVS3PmzKnRNUtKStS8eXMVFxcrKysrmuLWi/8umKSStcvNj5eV/hPUz/tfneDdrGKjiWWsaHNPqUqMJua/IwkclxW0j91xofsVG020Ta3UTrvN/WJ5vdD9Dck8LrB/YHssrxc45jsjW7meXeZxJUYTbTWy1cGzK+bXq8s163K92tRrQ7telqc0qnsjFtcL/F6T+8PJ6/3Xn6e+3nVK9fjlNyq6Kn7p76g9RhOd4N3coJ5f4PlIqrbsNblesdFE7T27leLREetli9FaHT0/RDxfsdFEKfKrvfdH81z7jXSt8Xet030b6fEjlSea+nTT9SLt6/TnppPXi9X9UNPnVpN7IVb3euA4SVE9ts1ope7erea96Dek741WUdWJG+6FuryevjOyLddr3m2ATrjsvojlSXah7XJJ6uDZJSm6125NHqvvz2gny1mfbcGG9Jp1Og6NKhgvLy9X48aNtWjRIg0bNszcPmrUKO3Zs0evvvpq2DEdO3ZUQUGBbr75ZnPb1KlTtXjxYn322Wc1um4iB+Obdpfqlb/cpIK0RfEuiqnUn6Em3rJ4FwNISPV9fyT79QAAieXVw6er+WkXqXmjtHgXpd7tOXBIxf96QcNSP4p3UWzxGW1v1uELNfzmh5SXgBlyp+PQqLqp7969Wz6fTzk5OZbtOTk5Wrt2re0xO3bssN1/x44dEa9TVlamsrKqF2pJSUk0xaxXL6zaqjm+4UrVYd2UtjjexZEkbnKgGvV9fyT79QAAieU3qR9J/07MYLReRD0It/7wGR1u5qEL9YhvuA6t2qqJQ7rFuzj1LiFfroWFhZo+fXq8i1Ej3/10QIakx3y/0bjUV+X1WDsaGIYsEzcF/h2Nmh4Xup/T1wvdv76uJ0l+o+KA0Pp26np1uWZd66W2f4+Gcr1ozxGr11lt76tYXq+m22J1vZocV9vr1fRctd0n0nFHOt8hI0Up8sf8vq1tebhedPs25OvF6n6o6TG1/ZvW9nq1eSwWdeKGe6G2dec3PPq3cYwkKdXrUUZqSvUHJKGywz4drhyzfIpnQ8T3/li+rmvCic9oJ8tZX23Bw4ZXD/uGS6qIqdwoqmC8VatWSklJ0c6dOy3bd+7cqTZt2tge06ZNm6j2l6TJkyeroKDA/L2kpES5ubnRFLXedGjZSB5J16Qssb3hg1+YtXmRRnNc6H5OXy94f5/hUUotAuPaXE+qXRBel+vV5Zp1+TvUpl4b0vVqc45Yva5re1/F8no13Rar69XkuNper6bnqu0+kY470vnSPL4jX7AW561tebhedPs25OvF6n6o6TG1/ZvW9nq1eSwWdeKGe6G2def1GFp+uKce8Q3XdQO6ujLLeP/StZqzfKPGpbys3mnrI+4Xy9d1TTjxGe1UOeuzLZjq8evGlJf1iG+4OrRsVLuTNHBRLW2Wnp6u3r17q6ioyNzm9/tVVFSkfv362R7Tr18/y/6S9Pbbb0fcX5IyMjKUlZVl+UlUF/XJ1Y0pL+uWBBozvsffuN6vWdtAHNWr73p1w9+xvu+PZL8eACBx3JK2SDemvqwRfRIzieW0RGyXB2sIn9H13RZ0+2s26m7qBQUFGjVqlPr06aO+fftq9uzZKi0t1ZgxYyRJI0eOVPv27VVYWChJGj9+vPr376+ZM2fq3HPP1d///netWrVKTz75ZGyfSZx0btVEg3tka+Xa7ok1m7qf2dTrer26XLOhzW5e39eL+2zqR7g/nLye3ey4NZkxNxGfX03LHuuZhGsyg7QT922yz+jMbOrMpn6k6wXaOdE8Fos6ccO9EMvZ1Ad3z07IibDqg127XEqw2dTr8BntZDnrsy3Ia7ZK1MH4iBEjtGvXLk2ZMkU7duxQr169tHTpUnOSti1btsjrrUq4n3766Xr++ed155136vbbb9exxx6rxYsXJ9Ua4ydcdp827y7Vk+9/oxUbduun/eWar4p1IRunp+qQzy9DUnqKV/vLD2tv2WEZfkOWdwkb3sq1+xqnp1Z7XOh+gTUMj3Rcba8Xun9gTcKSA4e0p3JNQieuJ0nelIp1CU+uXPP7s8o1v526Xl2uWevr1bJeE/V61b0+o7k36vr8or0/Ynm95plp9q+fAE/DfX7B6+BWd64aXy/492rq5Ujn86ZEeDz0GjV4fpbz1rI8XO/IxwXv6/TnppPXi+n9UIPnVtu/aa3rsraPGbU/ri7PP+rnVw/Xs329GfbnrtHrKaVqzeZrf9HFtUFNgF27vE6v3eoeq+/PaIfLWW9tT16zpqjXGY+HRF7aDAAAAACQfJyOQ6MaMw4AAAAAAOqOYBwAAAAAgHqWkOuMhwr0pC8pKYlzSQAAAAAAbhCIP50a2d0ggvG9e/dKUsKuNQ4AAAAASE579+5V8+bNY37eBjGBm9/v17Zt29SsWTN5aruqPICkVlJSotzcXG3dupWJHgG4Fu+FABC790LDMLR37161a9fOsmJYrDSIzLjX61WHDh3iXQwADUBWVhYNUACux3shAMTmvdCJjHgAE7gBAAAAAFDPCMYBAAAAAKhnBOMAkkJGRoamTp2qjIyMeBcFAOKG90IAaDjvhQ1iAjcAAAAAAJIJmXEAAAAAAOoZwTgAAAAAAPWMYBwAAAAAgHpGMA4AAAAAQD0jGAfgiEcffVR5eXnKzMxUfn6+PvnkE8vjGzdu1G9/+1tlZ2crKytLF110kXbu3FntOd9//30NHTpU7dq1k8fj0eLFiy2PHzp0SBMnTtRJJ52kJk2aqF27dho5cqS2bdtW43J/+OGHSk1NVa9evSzbp02bJo/HY/np1q1bjc8LwJ0a0nvh8uXLw97nPB6PduzYEdVzAoBQyfZeGKt2IcE4gJhbuHChCgoKNHXqVK1evVo9e/bU4MGD9cMPP0iSSktLdc4558jj8ejdd9/Vhx9+qPLycg0dOlR+vz/ieUtLS9WzZ089+uijto/v379fq1ev1l133aXVq1fr5Zdf1rp163T++efXqNx79uzRyJEj9ctf/tL28RNOOEHbt283f1asWFGj8wJwp4b6Xrhu3TrLe13r1q1r/JwAIFQyvhdKMWoXGgAQY3379jXGjh1r/u7z+Yx27doZhYWFhmEYxptvvml4vV6juLjY3GfPnj2Gx+Mx3n777RpdQ5LxyiuvHHG/Tz75xJBkfPvtt0fcd8SIEcadd95pTJ061ejZs6flMbttAFCdhvZeuGzZMkOS8dNPP0Xc50jPCQBCJeN7YazahWTGAcRUeXm5Pv30Uw0aNMjc5vV6NWjQIK1cuVKSVFZWJo/Ho4yMDHOfzMxMeb3emGebi4uL5fF41KJFC3PbgAEDNHr0aMt+zzzzjL755htNnTo14rnWr1+vdu3aqUuXLrrsssu0ZcuWmJYVQPJoqO+FktSrVy+1bdtWZ599tj788MOonhMABEvG98KAWLQLCcYBxNTu3bvl8/mUk5Nj2Z6Tk2OOtfnZz36mJk2aaOLEidq/f79KS0t16623yufzafv27TEry8GDBzVx4kRdcsklysrKMrd37NhRbdu2NX9fv369Jk2apOeee06pqam258rPz9e8efO0dOlSPf7449q0aZPOPPNM7d27N2blBZA8GuJ7Ydu2bTVnzhy99NJLeumll5Sbm6sBAwZo9erVNX5OABAsGd8Lpdi1C+1bnQDgoOzsbL344ou6/vrr9dBDD8nr9eqSSy7RqaeeKq83Nt8RHjp0SBdddJEMw9Djjz9ueezZZ581/+3z+XTppZdq+vTpOu644yKe71e/+pX575NPPln5+fnq1KmTXnjhBV111VUxKTMAd0mk90JJOv7443X88cebv59++unauHGj/vKXv+hvf/tbTMoDAKEa4nthrNqFBOMAYqpVq1ZKSUkJmwFz586datOmjfn7Oeeco40bN2r37t1KTU1VixYt1KZNG3Xp0qXOZQi84X777bd69913Ld9+htq7d69WrVqlf//73xo3bpwkye/3yzAMpaam6q233tJZZ50VdlyLFi103HHHacOGDXUuL4Dk09DeCyPp27ev2U20ps8JAAKS8b3QTm3bhXRTBxBT6enp6t27t4qKisxtfr9fRUVF6tevX9j+rVq1UosWLfTuu+/qhx9+qPEMl5EE3nDXr1+vd955R0cffXS1+2dlZemLL77QmjVrzJ/rrrtOxx9/vNasWaP8/Hzb4/bt26eNGzdaujUBQEBDey+MZM2aNeb7XLTPCQCS8b3QTm3bhWTGAcRcQUGBRo0apT59+qhv376aPXu2SktLNWbMGHOfZ555Rt27d1d2drZWrlyp8ePHa8KECZZuQaH27dtn+cZx06ZNWrNmjY466ih17NhRhw4d0oUXXqjVq1fr9ddfl8/nM8cjHXXUUUpPT5ckjRw5Uu3bt1dhYaG8Xq9OPPFEy3Vat26tzMxMy/Zbb71VQ4cOVadOnbRt2zZNnTpVKSkpuuSSS2JSZwCST0N6L5Sk2bNnq3PnzjrhhBN08OBBPf3003r33Xf11ltvRfWcACBYMr4XxqxdWOf52AHAxsMPP2x07NjRSE9PN/r27Wt8/PHHlscnTpxo5OTkGGlpacaxxx5rzJw50/D7/dWeM7DUROjPqFGjDMMwjE2bNtk+LslYtmyZeZ7+/fubx9ixW65ixIgRRtu2bY309HSjffv2xogRI4wNGzZEUyUAXKghvRfef//9RteuXY3MzEzjqKOOMgYMGGC8++67UT8nAAiVbO+FsWoXegzDMKIL3wEAAAAAQF0wZhwAAAAAgHpGMA4AAAAAQD0jGAcAAAAAoJ4RjAMAAAAAUM8IxgEAAAAAqGcE4wAAAAAA1DOCcQAAAAAA6hnBOAAAAAAA9YxgHAAAAACAekYwDgAAAABAPSMYBwAAAACgnhGMAwAAAABQz/4/rKkvxlxpzj8AAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 1200x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# [donotremove]\n",
    "plot_results(\n",
    "    (true_outlier[1], predicted_outlier[1]),\n",
    "    (true_cp[1], predicted_cp[1]),\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluation (metrics calculation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "\n",
    "pickle.dump(\n",
    "    predicted_outlier,\n",
    "    open(f\"../results/results-{model.__class__.__name__}.pkl\", \"wb\"),\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Binary classification (outlier detection) metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "False Alarm Rate 29.96 %\n",
      "Missing Alarm Rate 25.92 %\n",
      "F1 metric 0.74\n"
     ]
    }
   ],
   "source": [
    "# [donotremove]\n",
    "# binary classification metrics calculation\n",
    "binary = chp_score(true_outlier, predicted_outlier, metric=\"binary\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Changepoint detection metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Amount of true anomalies 127\n",
      "A number of missed CPs = 88\n",
      "A number of FPs = 69\n",
      "Average time 0 days 00:00:17\n"
     ]
    }
   ],
   "source": [
    "# [donotremove]\n",
    "# average detection delay metric calculation\n",
    "add = chp_score(\n",
    "    true_cp,\n",
    "    predicted_cp,\n",
    "    metric=\"average_time\",\n",
    "    window_width=\"60s\",\n",
    "    anomaly_window_destination=\"righter\",\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Standard  -  23.51\n",
      "LowFP  -  20.11\n",
      "LowFN  -  25.91\n"
     ]
    }
   ],
   "source": [
    "# [donotremove]\n",
    "# nab metric calculation\n",
    "nab = chp_score(\n",
    "    true_cp,\n",
    "    predicted_cp,\n",
    "    metric=\"nab\",\n",
    "    window_width=\"60s\",\n",
    "    anomaly_window_destination=\"righter\",\n",
    ")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "",
   "language": "python",
   "name": ""
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}