stationary_dgp_rl.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import statsmodels.api as sm\n",
    "from statsmodels.tsa.stattools import adfuller\n",
    "\n",
    "import gymnasium as gym\n",
    "import gym_trading_env\n",
    "\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "import torch.nn.functional as F\n",
    "\n",
    "from RLTrade.agent import DQNAgent, XGBoostAgent\n",
    "from RLTrade.utils import (\n",
    "    stationaryDGP, \n",
    "    nonstationaryDGP, \n",
    "    # build_features, \n",
    "    # build_rolling_feature\n",
    "    FeatureEngineering\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjUAAAGdCAYAAADqsoKGAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/SrBM8AAAACXBIWXMAAA9hAAAPYQGoP6dpAAC6mklEQVR4nOydd3gUVReHf5ueQAolpEDovYXQe5FQQhEFKdLrp2BDVASxgCIgKiCgqAjSBZQivUrvBEIvARJCGiEQUknb3e+PYXdndmdmZ7Ymm/Py5GHnzr137s7OzD1z7ikKtVqtBkEQBEEQRDHHyd4DIAiCIAiCsAQk1BAEQRAE4RCQUEMQBEEQhENAQg1BEARBEA4BCTUEQRAEQTgEJNQQBEEQBOEQkFBDEARBEIRDQEINQRAEQRAOgYu9B2BLVCoVEhMT4e3tDYVCYe/hEARBEAQhAbVajczMTAQHB8PJSVgfU6KEmsTERISEhNh7GARBEARBmMCjR49QqVIlwf0lSqjx9vYGwJwUHx8fO4+GIAiCIAgpZGRkICQkRDuPC1GihBrNkpOPjw8JNQRBEARRzDBmOlKihBqCIEoeKpUaR+6kILdAher+pVAviF5oCMJRIaGGIAiH5p9L8Zj6z1Xtduy83nYcDUEQ1oRcugmCcGgO3HjM2Vaq1HYaCUEQ1oY0NXoolUoUFBTYexhWwdnZGS4uLuTOTpQosvIK9LYL4evpaqfREARhTUioYZGVlYX4+Hio1Y77Jufl5YWgoCC4ubnZeygEYTHSsvPh5+XKK7CfffCMs52ZW0BCDUE4KCTUvESpVCI+Ph5eXl7w9/d3OG2GWq1Gfn4+njx5gpiYGNSqVUs0gBFBFAeO3knB4sPRuBT3HAE+7jj3WThn//WEdIM2H26Kwog2VVEv0Bu1AsTdQwmCKF6QUPOSgoICqNVq+Pv7w9PT097DsQqenp5wdXXFw4cPkZ+fDw8PD3sPiSBMJjY1G6P/vKDdfpyRh43n43D3cRaaVy2DXo2C8N2+2wbtLsSm4UJsGtMHGQ0ThENBQo0ejqah0Ye0M4SjMGbVBYOyaVuvAQBWnorB0FaVcSI61dbDIgjCjtAMRxBEsUOtViMmNVu0zoZzcUb7IU8ognAsSKghCKLYkZqVL7tNt/oBBmUZLxzT05EgSiok1BAEUezo8sNRWfV3vdceS94MMyh/liNfOCIIouhCQg1BEMWOrLxCyXXbVC+HhhV94eHqjG/6NeDsuxz33MIjMx0VLYURhNmQUEMQRLFmeOvKmNe/EUa2qWKwb1pEXawa20K7PbRVFXSs7a/d/vjvK0UiLlXi8xeo/tkefL3zpr2HQhDFGhJqBFCr1cjJL7TLn9SH7JMnTxAYGIg5c+Zoy86dOwc3NzccOHDAWqeGIGzKi3wlZ/tGoi72TJ0Ab8x+rRGGtKyMfk2CDdq+3akG3F2ctdvOTgqsGduSUyc+7YWFRyyf3otPAGC8tvILVXYeDVESePQsByNWnMPROyn2HopFIZduAV4UKFH/y/12OfbNr3vAy834T+Pv74+VK1fitddeQ/fu3VG3bl0MHz4ckyZNQvfu3W0wUoKwHo+e5aDD/CMAgKk962BS55oAgI82X9HW2f1+e+3nZlXKYvnI5qhW3gs1K4gH1asdUBp3H2cBADrMP2L3eDVpOTqD5dqf70XM3F4OH16CsC+zd9/EiehUnIhOxf86VsdnverZe0gWweKamuPHj6Nv374IDg6GQqHA9u3btfsKCgrw6aefolGjRihVqhSCg4MxcuRIJCYmiva5atUqKBQKg7/c3FxLD7/Y0atXL0yYMAHDhg3D22+/DQ8PD8ybN8/ewyIIs9EINAAwf98d7efbyZnazy7O3EdYt/oBRgUaANjxbnujdezJjcQMew+BMJECpQpPMvPsPQyjJKXr5s/fjz9A1Wm77Tgay2FxTU12djZCQ0MxZswYDBgwgLMvJycHly5dwhdffIHQ0FCkpaVh8uTJePXVV3Hx4kXRfn18fHDnzh1OmTUj4nq6OuPm1z2s1r+xY8vhhx9+QMOGDbF582ZcvHiRIgUTAIB7KZnwL+0BXy/Kc6SPh6szmlUpg8iHafYeCu8Yvvz3OrZOameH0RDmUmvGXgDAtkltEVa5jE2OqVKpcSs5A85OCtQJ8ObV8sWkZiP2aTa61KkAALgab5hCxBGwuFATERGBiIgI3n2+vr44ePAgp2zJkiVo2bIl4uLiULlyZcF+FQoFAgMDLTpWMRQKhaQloKLAgwcPkJiYCJVKhYcPH6Jx48b2HhJhZ47eSdGmELD30oop8NmVLT4cjbxCnX3N+vGtzDrGsuFN0fLbwwCYt2tXZ/uYGA5Ydtqg7FIR8soiTOPdDZdxatorNjlW9c/2aD+Pa18NX/Spb1DntZ9PIf1FAcqVckOlMvypgHLyC4vNvCeE3Q2F09PToVAo4OfnJ1ovKysLVapUQaVKldCnTx9cvnzZaN95eXnIyMjg/Dka+fn5GDZsGAYPHozZs2dj3LhxePz4sb2HRdgRpUqtlxOJUTPfTs7AvZQsew1LFrkFhsayCw7exc9H7mu3m4T4mXWMcqXctZ9n7byh/Txv7218s8t0L6RCpQoPnmThyO0UpL8owOaLj3D4lvx7sih4ZRHyYP9mCc/FDdCH/3EOVaftxqNnObKPs+JkDKpO243LcWkGUbFXnIzRfo5Py8HaM7HILVAi/WWgyafZ+bgioKV5ll384zbZVSTLzc3FtGnTMHToUPj4+AjWq1u3LlatWoVGjRohIyMDP/30E9q1a4crV66gVq1agu3mzp2LWbNmWWPoRYYZM2YgPT0dixcvRunSpbF3716MGzcOu3btsvfQCDsxfjU3J1KrOYdxYUY4ei5iPGxuf9MTHjKXOG3NjisJRuuUcjfv8eXspFPRrzsbh1mvNkRS+gv8eowRnGpWKI03Wwprj4WYsOYijtx5YlAu1/g3M68QPh60dFgciHz4DLeTM9G+ZnltWblSbihUqgzsvgAme/zJe0xesg7zj2BQ80qY2LkmqpUvZfRYuQVKrdD9+i+nEewrbG7Q/jvGLu2Lf28I1rkzuydCZx1AboEKM7Zdx4QO1dG+VnnB+kUdu2lqCgoKMGTIEKhUKvzyyy+idVu3bo3hw4cjNDQUHTp0wObNm1G7dm0sWbJEtN306dORnp6u/Xv06JElv4LdOXr0KBYtWoS1a9fCx8cHTk5OWLt2LU6ePIlly5bZe3iEneCbUKe/TPQIANcS0vHHiQe4nZyBzNyikSZAP/Bcdp5SoKb1iH2ajX3Xk7Xb7HMmB77zDwDnY55h8eFoZLMCB6az0jRsfqsNx47v9D1KxllcGLDsDGZsu46p/1zVlj3NzkfNGXux9VK8QX22ZhAANl+Mlxwl+7Nt3OsyMd3QYUaO9sfdxVmrGT129wmGrziHrLxCHLmdgg83RRWZZ4RU7CLUFBQUYNCgQYiJicHBgwdFtTR8ODk5oUWLFoiOjhat5+7uDh8fH86fI9G5c2cUFBSgfXudJ0flypXx/PlzTJw40Y4jI4oah1jLHwN/PYPZu2+h56ITaDTzAP448cDkfg/feoxBv55BfJp8FbqGqtN2o/pne3A9QacSV71U4/PFngGAz3tbxv20b6iu/5ErzhvkgsotEBeuPv77CqpO243ULOPeLoN/P4sFB++i/y+nceXRcxQqVXjOStNQyt2ZY8/w9rpLUr8GUUQ4F/PMoGwKKwQBAOQXqnAhlt9AnW0zJsTWS8a1mGzPQTHmvN6It/zUvVSMWXUB2y4nYMl/95i6e25h7dmHkvq1JzYXajQCTXR0NA4dOoRy5crJ7kOtViMqKgpBQUFWGCFBGEetVuPXY/dxIdbwIWZPTMk6PXv3LZOOpVKpMW71RZyPfYYxf14w3oCHu4917tl9lpzUfj4RzWgpvNyccWhKJ4N24ztUN+l4+jizVoMSnr9AvpJ7/jRB8TSo1Wp88vcV/BvFTCz/RDJv4XNenkMpk9Kdx5no9/Mp1JyxF52+P6otrx/kWC9dxYGY1GyTg8+pVGoUKlVYdSrGaN2hy8+i6rTdiE/LQe3P9wrWq/P5Pou5Vp++b6jpc3FScAzsywh4Rr61NlL7WePu/fvxB/hi+3WtjV5RxeJCTVZWFqKiohAVFQUAiImJQVRUFOLi4lBYWIg33ngDFy9exPr166FUKpGcnIzk5GTk5+veWEaOHInp06drt2fNmoX9+/fjwYMHiIqKwrhx4xAVFYW3337b0sMnCEn8fTEe8/bexsBfzyAlI9ckYcIa/Hzknvbz251qoFwpN0ntTDFKfc7SakSnZMnKxwQwBrWbLnCXhPddT8LJ6FQcu8ss4Ry/m4qaFUrjxizrhFfQt8vZfJE7nvtPsjnbg387i78j4/HBxihOec7LqMeZufLOARt9e5sgEVsJwjJ0+eEoRv95AQsO3pXddtBvZxC+4BhmSkhtcfr+UwA6GxdbMHT5OYOyQpUa7WqWx3uv1ESHWuXRtR6TuX61XpRtMTaeL9pmHBYXai5evIiwsDCEhTEZcadMmYKwsDB8+eWXiI+Px44dOxAfH48mTZogKChI+3f6tM6tMS4uDklJSdrt58+f43//+x/q1auH7t27IyEhAcePH0fLltJ/CIKwJOzJr+Wcw+goUd1rbVadjtV+/qRHHckajZjUbOOV9Ji4LpKz/cFflyWv5RcoVei+6DjHUwNgllyGr9A9jN1dmEdUKXcXTI+oCwDY8a7l4re835XraMDn/cEW+M6zNHMabQ0A7LvB2OJolqF8PFw4RqNyWPwym3iVcl4mtSekwV4yXHxY3JSBTXJ6LsavvoCLD9MQ+9T0ZVcxpGj8Fr8ZhjbVuSsdX+slbNWnaWU/AMBH3etg7bhWcHt5f3Wq7Y/OdfxFWupYfuIBCpQqo0uz9sLiQk3nzp2hVqsN/latWoWqVavy7lOr1ejcubO2j6NHj2LVqlXa7YULF+Lhw4fIy8tDSkoK9u/fjzZt2lh66AQhidwCJS7qBUwz5r5pKwY1D9F+dnZSaG1TjPFCxgNKM8nr2w8cvp2CDvOPiL71Hrr5GK/8cBRX49Px4IlxQer7gbqYS291qoHYeb3RuJKf5LEaI8DHA5e+6CZaJ08gF5O+tgYAkp4zqvmKZbzw87CmksfBth3yfqk9OvugaC1tFlembbmKqtN241YSN6THhnNxsvpRq9V4npOPaVuv4tAt8SWrFaOa4+SnXUTruDor8N9HnXDx83CDfcdeGpvzaVA1gkjjir7463+t0a4mI9j0b1oRjSr6ih5z0eAwwX2rxkhTEmTlFaLvkpPoOP9IkRRs7B6npqjh6LEhHP372YK/zvM/DIvCDa7RbAxsVgkA8EazSvBwdUKgjwcCfNwF20ldNjl86zGqTd+DBQfvYky7qrx19N96n+fkY8SKc3j0LAfj11zEg9Rs3oBzfFQvX1pSPXPw9TS0K1g7TveAT8sxHrujbiCTmiExnRFug3094Ovpipi5vTA5XDjshIaGwbrJqLSHbkmsQEnJLc1BqVJj48slzoifTmDd2YfaZ+BTCcbdOfmFuJfC2H19tu06mnx9EEcFvNs+DK+NmX3r488xLdC1XgAqlfHCkY87C/Z9aEonVPcvjfKlDe/L/62NxH+3H6P13MOYwfJ2KlSqtAlPNdfJ+vGtceyTzvj+jVBRoaZ8aTdUNqL909dcCnE7ORMpmXlFMu4VCTUvcXZm4nawbXsckZwcRl3q6krxL0xllsAa+rUE+4cd/+mlQPFvFJNPTaOJOD3tFWzTC7v/7zu67SG/nzXIhs3HuNVMOpPFh6Mlr603+fogTkSnSvbIYOPDI3BYGna8Gg3Nq5TVfv5+3x2D/fpo8v3M2HYdABD40h5GoVCgZdWyYk0BAJ1Yqv9mrND6pgRmIxiiHj3HpPXcJdLPt1/H6pdLtKvPiHvy5Beq0GPRcYQvOI6zD54KvsxoGNO+Kka3q6ZNQwAAlcsKCxFi+wBg7KqLeJyRh/UvNUrvbriErguOAWCu2bJeOnu5KuVKwdlJARdnJ0yPqIvypd3xv47VMbVnHXzZpz4aVvTBvskdRY8HAKPbVjVah02fJSeLjD2hhuIdD9mCuLi4wMvLC0+ePIGrqyucnBxL3lOr1cjJyUFKSgr8/Py0Qhwhj0KRN+fBv53Bg7lFIyVBPmucGjfhMqyH4FsdqyNULyJvvS/3yUqpILZkJRR0TIgrX3ZH6NcHOGUj21ThFThsgaeb7v7YejkBCwY3QYqI10f6i0L8b60uf911VkLKMjzG2nUDvfHXhNYI++YgqpTzQk1/nUbKyUmBBsE+uJGYgahHz1Hd3/raKkcjO68Qr/18inffzJ03eY17k9NztcLoyJXncfyuTiMzd49xD0G+QIlC1++kzjU4huF3Z0fg/pMsDFh2Wmt0zuZ6Qjp2XdXZmfp4uMBJoO+3OtXAW51qcMrGtq9mdPwAUFbEsWDLxLa8GtZdVxPRr0lFSf3bAhJqXqJQKBAUFISYmBg8fFj0ffFNxc/Pz6Y5tByNmjOE3TFVaiDuaQ4CfN3h7mJ7oZFtuPpVX8PcLx6uOiFD4/VgCSqX9UKcnkZBc54OTTH+dggAPp6Gj6KZfcWNHq3FeJ4J4NGzHFFNU2pWHsfwtDorMqxmaSrY10MbKE2lVqNMKTdBIVKTpXvK5ivoWNsfzWcfAgC82bIyZvSuh9I80ZTTsvPx1tpIBPt5YNEQYdsJRyY9pwD/3XmMDzddMV5Zj7uPM7VCDVugAYTtqjTM5LnfxNDXbbi5OKFekA+cBSJOs8MdAEBajvUC4n3Vt75WG/3dgEb4fv9dHJ/aWfCZ9sHGKChVaihVagxk2fTZCxJqWLi5uaFWrVoOuwTl6upKGhoL0qWOv0H02I7fMxOftZJInr6XioTnL3gfHmzDVS83w99ZoVDg3S41EZ2SieZVmCWOGv6lOG7LarVaMJT/lUfPecs/7VkX72zgDxQXvuC40FfR8uPAUINjdqhVXvBN1Br8MDAUH/99BcG+Hpjak/GyCq8XoA1aqC/QNK9SBt4eLvjmtYa8brrz39AZOCsUCu31sPJkDL7ffwdz+4snnS1f2l0rJGkEGoCx5/JwdcJXegJfboESYd/okgU7Oznhx0GhRr+3o/HplqtaTzS5jFx5HrHzevNqY/N5yjrUKo8ZvevhRb5SNBv3hgmtDNyrhUwbvdydkSkzNIKlGdmmKnLyleharwLqBvpgcAvjqUI0AQZrVCiNpjbKTC4ECTV6ODk5wcOD4kMQxvl+YCi2XorHnD23Dfb1+/kUUjPzcPSTzhbL/pyTX4ihfzAPxyrlSqFlNWFbDWeB5dOPe9ThbO+b3BF/X4zXhl7fdz0ZEY34g1r2E1Dnt69VHl/3awClSi1obyTElS+7w1cvAFjdQG/8PqK5rH7M5Y1mlfDGS+NqDctHNkO16XsM6taqUBr/TGwLQNiQV+g3H9u+Gka1rWp0WW3VmBYGb+carsWn41ZSBuoGekOhUPAGa9tyKb5ECjWmCjQaktNz4eVu+ELA56k3tGVl1A00HjCxbY3yuPdtBFycndDi20N4kpmH8HoVeOv+OrwZ3vvrMuLT7OdN6eykwDtdavLui5nbC4+evcDdx5kYv+aiwf7+v5zWfld74ViGIwRhRdgT2KEpnV4a49XAvP6GocavPHqOhOcvsPS/ewb7TOU8y4X69+P3RWrq4lEYw9XZCX1CdULMB5uiZI/L19MVI9tUxZh21WQFjPv77TYcgWbN2Jbo0zgIGya05ti02AshjRVbWOATXs7P6CrarxQ7oXoi0YUvPkxDxE8nsPtaktYThg+5LsuOzIM5vRD1pbjrPgC0nnsYvxwRv7deqVsBr4YGCwr/fGgm+f2TO2L7O+3QXMB4PKxyGZz89BXEzuuNcSJ2MBdmGLqB2wKFQoHK5bwQXj9AMFWJfsBKW0NCDUFIZNJ63RIL2w14cAvhdeS/L1ou+uZoVioCFxFD9iEtQmQZl3qxMnaLTZJ8HP+EG4uD7VElRP+wiujdOIjj5QMAHWv7Y+nQpqLGikUBtvs1ALzLeqv9a0JrVPA2X9Pr7KTQBuET4t0Nl3FRJE2HfuLDkka5Um64MasHYuf1hpOTAn5ehtfVv++0Q6RenBhNlnYhVo5uYfS3EaJsKTc00TPQF2JGL0OhIayyH2Ln9Ya/t3B4BlsxvkN1uPEI9b8ZOX/WhoQagpDIwZu6pJDsnClCb/QA11XXVPILVfh+P3eJy92Ve+uyJ7cx7aR5OmjQVxXzRTNlZ5b+gBXLIqSsJ6deBR8PTpJIfX4d3hQLBjfBz0Ob2tRmxlR2vdfeoEx/3B/3qIObXzOTZ5sa8nPZCdG3sXFNgGY5UojpW6+K7nck9rOWnj7vXQ+RX3QzSIPB5vY3PREa4odyPHFihGgQbLv8XE5OCtSswH05+WtCa5sdXwq73je8P7ZeTsCTTOMxgKwFCTUEYQL6gsDmt/gjXOcVyA+eplSp8cay05jwcs16zZlY/KynEtfEodHwkBWuXf9BKAW2IDKHJ8Flg6/2az8Pa10Zvw5vir/fbsMr0C0e0gRH9YKOPZjTCzdm9UDPhsUrCW3Dir4co29NNFd92Nm1LYWYsCxEaXcXjibgryKep8eSsJMwdqvP7903qk0VAIxg7sHSUEoVVhYNbmL6AE3g56HcqNTsMRcFagk8a+wZNJKEGoKQSZ0Ab4OyltXK4s8xLVBBTy185sFT2f0/eJKFiw/TcPDmY7zIV2InKz4FG3YCSXYALFNiu3zP8tbxcHXGjiuJeC4QSdfHwxU9GwahhYBdgEKhQFWWSzPAvHWKvTUXda7P6oFvXmsoyS7Dklz6ohuOf9IFsfN64+7sCKP1+zetiG2T2hqU5xYoHSaauFqtxjvrL+GnQ7rI1fraxSrlSuk3AwDM6tcQsfN648NutTnlPw1pYlC3RdUy6NEgANsmtUXH2v74cWAoavHc+9akTqA3wutVgL+3u6Tf39ZoPPti5/XGe6/olmGD/TxFWlkXEmoIQiK1A5i3kum96vLu71KnAs591hUfhteGy0vBIik9F/P2GnpHibHtsi7ezLqzD5EuIFw0ZGlPNBoEY7lfhPBwddYuJf12/AHe/+sy+v9yGrGp2Vj6X7RBXSnEzO2FrZPa4v6cXiaNqShR2t0FI1pXsYpGRoyypXSh7d1cnAxsmPT5uEcdKBQKvB6mC4Y2fvUFhM46gHf/umzVsdqK2btvYfe1JCw8dBcFShU2XYhDnc/3mdUnnxD0x8gW+G1Ec4RVLoM1Y1tigJ53nK34Y1QLXJgRLqglLCq0rm65pVdzKL6vTgRhI9Ky8/FvlG6d2FNkUlcoFPggvBa61PXHq0sZF+hfj93HxE41DFyXhfjlqG6p6e/IR6KZgC/FpaFp5TLa2BYVzXhDepHPVRk/SM1G5x+OcsrKSPwOAHMu7B2zwtGoXM4L2ya1xf4bjzkGrb0aBeKj7nW0UW2/7tdAKxxrki/uvpqE7wYU8gbuk0OBUmWxMAWmsJqVib7XTycQrZd/6K2O0jLTs3F1dsK+yR2gUgFuLgrkFqgk368EQ7ua5bFqTAuTlr8tSdEW/QjCCIdvPcaqUzHa7ec5+Xh7bSTWnIm12DEmrLmImTtvaqN4SnE3rq2nps7IlRYBVD/Xz93H4gnj+v/ChC3XaHP4IvNKJVVCgj8h2yHCdoRVLoNpEVxt4eIhYajB8njz5gnZDwBn78tfDtWQlp2PqtN2o9aMvThjRj/m0rOhLiK6vkADMMEgTaFuoA/qB/ugZgVvNDRR41nS6VynAiqVEc9pZW1IU0MUWwqVKm2CRTcXZwxsXglNvmaiqu67kYy07AJ8ICFDsjEuPkzjbBcojdsmuOupisXyJGn4fv9tA4NgNh90rYX2tcqjbCk3dP3xmLb8yJ0U/HDgLgDguRnh0yv6eSLhuXjQL76MwoT9kRrsTGmGXc3wFTpPqzeXn0XsvN44cjsFDYJ9UMGHcWNXq5kAjE2rlMGrIl5w5rBLwMZMQ3HwqiOsB2lqiGLLKdbb4mfbrqGWXl6mhYfuChq7no95hphU04JESfGU0PdcScs2nnpDTKABgA+71UaLqmU5b+QANxbO8xemCzVvtjSet4Uv2iphH8KMBFh8pa5h1NpcCcK1EDdYSToBxt5rzKoLaDnnsLZsxvbrWHU6Fu//dRkqK2RvNjb+FaNsG4maKHqQUGMl4tNyUG36bgz89bRoZmfCdM7HGFeBX0tINyhLfP4Cg347gy4/HOXEXxFC3wbBFLfKZxKEGjauzuJvmxv/p4tXUa6UTnvC9mKSy7j21dGncRB+GtIEc3miJE/oUM0uiToJfsJCxO2VFgwKRbhe4lJTQgxo0A8at/DgXe1nzbIpO4rxHycfmHwsIYSWcSt4u2PDhFYWTdRKFE9IqLES7b87ArUauBCbhhPRqfYejkMS5GvcKHbTBcM4HXeSM7WfG3y13+jbHzu7tZzQIdNZdg9PjQg16XoalhWjWnC2F+jl8WldvRyGt2YSzbGXjIRcWaXg6eaMpUObol+TinizZWXsZgXWer9rLczoLS8TMWFdShnRmvl5uWH5yGb4bkAjhFZibESkLIMKEaWX0JR9Ta8+HYvDtx5z9vPlRDPGtfh0vPfXZcQJGMdn5fK/hGyZ2BZta5SXfTzC8SChxgqc1YtNkp1v36yrjsqxu0+M1tGsv8c9zcGsnTew80oixqy6wKlT94t9eGc9f5bp1Kw8pGYxD+8udfyx74OOksf3VqcaGNSccQO9pGeXwya3QImmrAzLANBWLzJt/6aG7qRuLzOu/3eb8W6Rk3dJCg2CfbUxKKboxfUg7M/49tXRsKIPR3jWR6FQYHCLylpX9AM3TUv4aCzGzR8nY7T2bebQd+lJ7LySiI7fH4FarUZadj7WnonVLiOzo3oTBB8k1FiBIb+f5Wy/u8Ex4kMUNYQecHUCvLUGrRqVecfvj+DPU7F4TyBWx+5rSbwP7gmsTLRz+zdGnUB5wbeuxjPLX1svJwguQw767QwneN4vw5rCxdkJH4YzgkRvgcR5G84/5GwnpefKGhtRvPH1csWu9zrgrU41jNbVBIE8dU++19I76y/xZiuXwhU97Y4+KpVaqynVz2PVeOYB9F92Gl/8ewPvb4wCAMxlxXxiB5m0Z7A3omhBQo2FuRr/3N5DKNFMj6iLPR90wNf9GgBgVOaf/H1FUttq0/eg7dzDWjubeymZuBz3XLs/0ARNCNumodbne3mNJzWCDwAMb10ZvV4KMZO61MD68a04WaHZ5JphH0EQALD5wiPM2HaNdwk2K68QVaftxu5rXG+jJTKSOfb7+RTSsvORV6jEO+sv4XJcGh5n5OLmS6PjJl8fQN0v9uFeShYm6mlLM/MKtcb8x+8+waDfznD2X5vZHStHN8ftb3qaFEWbcExIqLEwmoBrhHXJ5DEYPPJxZ7zVqQacnRScbNN/R8ZL7jcxPRfXE9KhVqsxYU2k8QZGGNqqsvazWg1U/2yPQSwaNn6eukzCrs5OaFezvKBhsn4WXy8J8XOIkslXfQ3toa4npGPqlqtYfy4Odb/Yh3VnH+JFvhKRD59BpVJzstKzEUtYCgAD9JZKw745iFd+OIbd15Lw+i+n0WrOYfRafAIPnmQh46WNTPiCY0aTIJ6P0WlyvuhTH15uLnilbkCRy4dE2BcSaqyIHysi5b2UTJGahBzOxzzD0OW6mBmnpr2C+3N6oRor31ConqeGHAb/fhbdFh432eWbDZ9a/It/r2s/333MvS7kxNjQD9se+blt8xIRxYeudXUaw9m7bkKtVqPPkpOcOp9vv456X+7DgGVnMHv3LdTwFzY6vzO7J7rU8eeNYP0ljwDFF//oFVasJblojOQJQh8SaiyIvk3Guc+6aj+HLzhu6+E4LIN+O6N11a5SzgsV/TwN1M/Vyhv3Ajr4obDR7z29SKWnpr1iwkj5YWuRui/kXhf9mkgPWFa2lBv2ftABc/s3wuUvukmKdEyUTMqU0r1g/XEyBvefiAvsK0/FiKZCcHdxxp9jWuLkp104gs2m/7WGr6dhNOPKZS0bZZZCCxBCWFyoOX78OPr27Yvg4GAoFAps376ds1+tVmPmzJkIDg6Gp6cnOnfujBs3bhjtd8uWLahfvz7c3d1Rv359bNu2zdJDN5tIlodLsyplDG48cwJflQRUKjXOPngqmlJAX3AUW0sXE0RWjGqOWgHeiJ3XW5t8UogPutYyK6fSvW+52XWFQrAf/6SLQWA9Y9QL8sGbLSujTCk345WJEot+2oTXfzG+TM4O6qhBP6miQqHgvEC0epnU8NrM7px6Senikarl8JoMwZ8oeVhcqMnOzkZoaCiWLl3Ku3/+/PlYsGABli5digsXLiAwMBDdunVDZqbw8syZM2cwePBgjBgxAleuXMGIESMwaNAgnDt3TrCNPTh6R+divPilMR1bW7PxfJxBG0LH1ssJGPL7WQxdfpZ3f+LzF/h+/x1O2QORN059QaR19bKIndcb9+f04gTp+tCIu/LgFsYj7Yrh4uzEeXv9/Th/UDJNNmaCsAZsA99MgXgvbNJ4Um40qeRnUDYtoi5cnRV4p4vOC8vbwxVHPu6s3ZaSWkRD7Lze+O+jTlg3rpXBviBfDywaIt1QmSh5WFyoiYiIwOzZs9G/f3+DfWq1GosWLcKMGTPQv39/NGzYEKtXr0ZOTg42bNgg2OeiRYvQrVs3TJ8+HXXr1sX06dPRtWtXLFq0yNLDN4ulR+5pP2sm1AAfnceMnBu7pHEnORPf7WPcNa8nZPDW6fLDUU4Gayl0r68TXl5rUhGAoXZnXPtqWDasKX4cyO9lVNrD/BRp+okgO84/wlni+veddmYfgyDE0NeyaNj1Xnvecg2f9qyLf99ph16NArUva2waVvTFtZk98EkPbryckDKmazer+5dG4xBDjebu9zuY3CdRMrCpTU1MTAySk5PRvbtONenu7o5OnTrh9OnTgu3OnDnDaQMAPXr0EG0DAHl5ecjIyOD8WZoCpQpT/7mCqtN2a8uE7CK+3XPL4sd3BHLyC9Fj0XGj3g95hYYuzBM6VBNtU8FHl0KgjV5AOw0ers6IaBSEAc0qYWSbKgb7S7mZL9TUCfTmuHfHPctB+AKdoWTjSpQVmLAuQlnY6weJ5zJ7LSwYoSF++GVYM8GwBnweSFKTbLJpWFE3Fh8PV0R+Ho7Iz8MBAFXLeaEsLbMSRrBplu7kZCaaZUAANz9HQEAAHj58yNdE246vjaY/IebOnYtZs2aZOFppFCrV2HyR6zKsH6q+jJcrryqXYNBPlAcw4dLrB/tArVbj7INnqM7jifFZr7qY0KG6aN+9GgVh3Vlm2U9KCoH3u9ZC1XKlUCugNKqVLwVXZyeLxcD4aUgTNPhqP+8+/QSYBGFpnmUZpupoV7McnJwU+HV4M7y9jglhsHBwKD7cxMR2+qpvfUnpSMzh33faITolCwOaVjS4D8q9DKIZO6+3VcdAOA42FWo06F+4arXa6EPdlDbTp0/HlClTtNsZGRkICTHPPkIfPo8T/ciYr9QNwJZLjOBzLT4djeitnMNnW68ZlPVdyrib9m9aEVsvJRjsf6tTdfyvo/FIqm2ql8Pykc21uW+MUb60O8a2F9f+mEopdxcseTNMMKoxQViTtzvXwI+sJJQA8MPLJdeeDQM5gsPrYYZpOSzJiald0GH+EUztWQehIX5mhWAgCDY2FWoCAwMBMJqXoCBd6PeUlBQDTYx+O32tjLE2ALO05e7uLlrHEoxqUwWrz+g0TaPaVuXsT2TFaEjOyEUjkFCjQa1WI1rPfZoNn0ADAKE8Bot8KBQKdKtfdDL38nk+TRPJ3UMQlsLV2UkruChVaiggLy6SacdUGNgSvtWpOkLKepH2hbAKNrWpqVatGgIDA3HwoC55X35+Po4dO4a2bdsKtmvTpg2nDQAcOHBAtI0t+ax3PfiwjEl7NAjk7J/1MmQ/AFA0by5XWCkCpNKmerkiJajIoWo5L21eKg3GltAIwtI4OymsLtAAwNKhTbWfP+9dj1naCqfkqIT1sLimJisrC/fu6byAYmJiEBUVhbJly6Jy5cqYPHky5syZg1q1aqFWrVqYM2cOvLy8MHToUG2bkSNHomLFipg7dy4A4IMPPkDHjh3x3XffoV+/fvj3339x6NAhnDx50uD49sDdxRmXv+yOawnpvEZ3tQN0SRDJA4rLw6fyo/ZumNCq2NqgKBQKXPw8nGNYTnlrCEfFhxUfZ2y7ajYRpIiSjcU1NRcvXkRYWBjCwhjXvylTpiAsLAxffvklAGDq1KmYPHkyJk2ahObNmyMhIQEHDhyAt7du4o+Li0NSki6JWtu2bbFx40b8+eefaNy4MVatWoVNmzahVSvDOAb2wtlJgSYhfoJuk51q+wPgz1nkaOy7noyvd97kZJ4WwpQJvbgKNGx+fvkGO7addex3CKIoEFbZDxX9PNG5jj8JNIRNUKj1Q7Q6MBkZGfD19UV6ejp8fMTdGC3N+39dxo4riQCYZGzjrGSMWhTQaCGWvBlmNPkdW2Nx8+seyMwtxG/HHmDlqRjBNo6yFp+SkQt/b3eHENIIQgilSg0nhWO8jBD2Q+r8TbmfbAQ7gNs3u6RpMYoj7O8V/Vg8iaf+OfByc0GAj4dBQrz3X6mJ6RF14enqjM97czNTF2cq+HjQg55weJydFHSdEzaDhBobUUrP9TvWBFuS4sB/t1O0nxf/d0+kJvBUIBgYAHzSo47283tda+GtTjVw65ueGE9GtQRBEIQAJNTYiKhHzznbfZdwjZzTXxRg66V45OQbz8lSlMnniforRG6Bru6JqV04+97pUhPXZnbHvW8jRLMFEwRBEIQGmi1sRAVvbnjxnHxuxu7QWQcwZfMVfLgpyoajsjxPMnM52wOWnUZMarZBdm0AePEya7mTAggpa5jM0dvD1aRQ6wRBEETJhGYMGzG0VWWDGDUam5PIh7oIxPtvPLblsCzKk8w8zNx5k1MW+TANXX44iq6sPEcasvIYT7BgP+uGYScIgiBKBiTU2Ih2Ncvjylfd0ZaVVLH3yyWoAcvOaMtcnYufQd0rPx5F1Wm70eLbQ4J1HjzJxuFbjzkam8TnjFbH39v6UZ8JgiAIx4eEGhvi7eGKBYOaaLfzC1XYey2JU6dAqeZdqimqnLn/FA+eGBo9V+TRvoxbfRHVpu9B1Wm7oVartTmQLsc9t/YwCYIgiBIACTU2xtuDG8R54vpLBnUO3UoxKCuq7L2exFteP1g8DlDYNwdF9xMEQRCEXEiosTFePFm99YlNLT7u3rUqlDYoC6vsh/B6FUTbPc/RRVamqLoEQRCEJSChxsYoFAr8NKSJaJ1v99yyzWAsADs7uYYeDQIxqHkI/hzdAtsmGU862rxqGWsMjSAIgihhkFBjB/o1qWi0Tm6B0mgde3L3cSYiH6bhXkqWwb7hratAoVCgS90KqMzjqq1P3UBvo3UIgiAIwhgk1NiJV43kROq9+ISNRmIa3Rcex4Blp7XbPh4uWDuuJe59G4HS7jq7IXZ6iImda+DNlpUN+qrub7iERRAEQRByIaHGTiwc3ISzPaNXPZz7rKt2+1l2vo1HJI1CpQoPeVI8nPssHB1q+RsEy3N30dkQRTQMxLj2Va09RIIgCKKE4mK8CmENnFmR+LZNaouwyly7kjeaVdJ+TsvOx93HmWhZraxdE8Op1WrUnLGXd5+niAH08U+6ID4tB40r+SHh+QvOPspzRxAEQVgKEmrsyP7JHRGTmsURaCZ2roFlR++jQMnEqll8OBoLDt7V7v9zdAu0q1keTgrYPIXAipMxJrWrXM4LlcsxtjVBPtx0EZc+72b2uAiCIAgCIKHGrtQJ9EYdPSNZl5canFWnY/F573ocgQYAxqy6oP08f0BjvN60os0SPs7eze+VNf+NxpL7cHJS4P6cXjh2NwUBPh4oU8rNUsMjCIIgSjgk1BQxVp2O1X5ee9bQXZrN1C1XEfs0G1N71rXyqIT5eWhT9G4cJKuNs5MCr9QNsNKICIIgiJIKGQoXMb4boNN6bL+cYLT+L0fvG5RtPB+HqtN240ZiusXGxZe64ceBobIFGoIgCIKwFiTUFDF6NQqCr6crAOBKvHyh5Pfj9zFt6zUAQO/FJy02rtF/6pa9ejYIxKb/tcYAljEzQRAEQdgbEmqKIJ6uXE+i8qXdcX9OL966+sHt5uy5bZUxHbv7RPu5fa3yaFW9nEhtgiAIgrA9JNQUQbzcuUKNk4LrAs7G3cUJPx64g8txaShUqmQfS6lS41JcGl7kS49g7OZClw1BEARR9KDZqQgyuHkIZzslMw8A0LYGox15Pawitr7MqRSdkoUl/93D67+c5o0hM2HNRdFjvf/XZfT/5TTqfbkP68+JGyZrMBYNmSAIgiDsAXk/FUGS0nM525/3rgcA2DChNVIyclG+tDueZOVJ6uvgzcfIK1RyIvuy2X0tSft5xrbrGNaqimh/v41oBg9X45nGCYIgCMLWkKamCDK+QzXtZ39vd4xqW1W7XcHHA05OCpSTEd8lPafA7DGVfXm8quVKmd0XQRAEQVgDuwg1VatWhUKhMPh75513eOsfPXqUt/7t29YxirU3lcp4oUmIH1ycFDj/WVfe4Hpyogk/f6ETarLzCrWZtVUqQzftn4/c4+1DkzXcw5XkYIIgCKJoYpcZ6sKFC0hKStL+HTx4EAAwcOBA0XZ37tzhtKtVq5YthmsXtr/TDvfm9BLN9bR0aJjgvmszu2s/P88pQG6BEtcT0tHgq/0IX3AMb6+NxIsCQ+Pg7/ff4WzfSExH1Wm7kfPSkFgsxxNBEARB2BO72NT4+/tztufNm4caNWqgU6dOou0qVKgAPz8/K46seNGncTD6NA7G8uMP8O0eXQqD+kE+8PZwRZMQP0Q9eo5Bv50xaLvvRjIGx4YYlOujH+vGx8PV/IETBEEQhBWw+1pCfn4+1q1bh7FjxxrNQB0WFoagoCB07doVR44cMdp3Xl4eMjIyOH+OyISO1RE7rzd+GtIEtSqUxqIhTQAAfl7iAsjBm4+1n9keV4+e5QAA8gsNXcTJSJggCIIoqthdqNm+fTueP3+O0aNHC9YJCgrC77//ji1btmDr1q2oU6cOunbtiuPHj4v2PXfuXPj6+mr/QkKMayaKM/2aVMTBKZ1QO4BJklnGS9yYeMO5OABAkxA/DGqhiw6syTllyTQLBEEQBGFtFGq+pD42pEePHnBzc8POnTtltevbty8UCgV27NghWCcvLw95eTrX54yMDISEhCA9PR0+Pj4mj7m4MGfPLfx+/IGkujFze6Ha9D3a7e/faIxP/rlqUC92Xm+LjY8gCIIgpJCRkQFfX1+j87dd49Q8fPgQhw4dwtatW2W3bd26NdatWydax93dHe7u7qYOr9ijn25BiNbVy0KhUMDbwwWZuYUAwCvQEARBEERRxq7LT3/++ScqVKiA3r3lv/1fvnwZQUGUIVqMjFz++DRL3uR6TW0Y3xoAMLRlZcG+3u9aC4c/EjfkJgiCIAh7YjehRqVS4c8//8SoUaPg4sJVGE2fPh0jR47Ubi9atAjbt29HdHQ0bty4genTp2PLli149913bT3sYsUHXQ1d3sMq+6GvXpoDp5d5pXxFDIundKuNGv6lLTtAgiAIgrAgdlt+OnToEOLi4jB27FiDfUlJSYiLi9Nu5+fn4+OPP0ZCQgI8PT3RoEED7N69G7168WeuJhj8vNxw5cvueJKVi/AFjFH1tkntAAD7JndAz0UnMH9AY2396uX5hZZejQKtP1iCIAiCMBO7GwrbEqmGRo7IgydZKO3uggo+HoJ1VCo1fjhwB78cva8taxLih99HNkMFb+F2BEEQBGFNpM7fdnfpJmxDdf/SogINwCxDTe1Zl1P214TWJNAQBEEQxQISaggD2DEQKdcTQRAEUVygGYswoGvdAO1nY1GeCYIgCKKoQEINYQBpZwiCIIjiCM1ehAGdajMJR12cSEtDEARBFB/sGlGYKJoMaFoJCoUCLaqWsfdQCIIgCEIyJNQQBjg5KfBGs0rGKxIEQRBEEYKWnwiCIAiCcAhKlKZGE2cwIyPDziMhCIIgCEIqmnnbWLzgEiXUZGZmAgBCQkLsPBKCIAiCIOSSmZkJX19fwf0lKk2CSqVCYmIivL29LRp/JSMjAyEhIXj06FGJS79gS+g82w4617aBzrNtoPNsG6x5ntVqNTIzMxEcHAwnJ2HLmRKlqXFyckKlStYzgPXx8aEbxgbQebYddK5tA51n20Dn2TZY6zyLaWg0kKEwQRAEQRAOAQk1BEEQBEE4BCTUWAB3d3d89dVXcHd3t/dQHBo6z7aDzrVtoPNsG+g824aicJ5LlKEwQRAEQRCOC2lqCIIgCIJwCEioIQiCIAjCISChhiAIgiAIh4CEGoIgCIIgHAISagiCIAiCcAhIqCEIgiAIwiEgoYYgCIIgCIeAhBqCIAiCIBwCEmoIgiAIgnAISKghCIIgCMIhIKGGIAiCIAiHgIQagiAIgiAcAhJqCIIgCIJwCEioIQiCIAjCIXCx9wBsiUqlQmJiIry9vaFQKOw9HIIgCIIgJKBWq5GZmYng4GA4OQnrY0qUUJOYmIiQkBB7D4MgCIIgCBN49OgRKlWqJLi/RAk13t7eAJiT4uPjY+fREARBEAQhhYyMDISEhGjncSFKlFCjWXLy8fEhoYYgCIIgihnGTEfIUJggCIIgCIeAhBqCIByXWzuBhY2ARxfsPRKCIGwACTUEQTgum4YD6XHAhoHAvcNA9lN7j4ggCCtSomxqpKBSqZCfn2/vYVgFV1dXODs723sYBGF7XqQB6/oDvpWBD6/ZezQEQVgJEmpY5OfnIyYmBiqVyt5DsRp+fn4IDAykOD2E45OZbFiWHmf7cRAEYTNIqHmJWq1GUlISnJ2dERISIhrcpziiVquRk5ODlJQUAEBQUJCdR0QQFiA7FfAsC/Ddrxf+sP14CIKwKyTUvKSwsBA5OTkIDg6Gl5eXvYdjFTw9PQEAKSkpqFChAi1FEcWbhEhg+StA6JvA679y993eDRz/Xrht0lXAswzgR8E4CcKRcCx1hBkolUoAgJubm51HYl00AltBQYGdR0IQZnLspdBy5S/DfRuHCrd7Hgf81gFY1NA64yIIwm6QpkYPR7c1cfTvR5QQUm4Bd/fqttMTgBM/ABWbAbd2ibfd+6l1x0YQxYnCfODOHqBSC8C3or1HYzYk1BAEUfy4uJK7vbA+fzkfd/boPkcfAkJaAh4UYZwooRybB5z4kfk8PQFwL23f8ZgJLT8RBFH8UFoo7ML6AcC6AZbpiyCKC7f3AIsaAXHndAINAMQX/yCVpKkhCKL44V/Xcn3Fn7dcXwRRHNj4JvP/yu7ccrXS9mOxMKSpIQiiaPPoAnD4a6AgV1emJEN3grA4hcU/8CwJNcWYNWvWoFy5csjLy+OUDxgwACNHjrTTqAjCwqwIZ1Tkuz4E1Gqm7EWa9PYz04FRRoyH2RTmA/f/A/Kz5Y3TXFQqQFlo22MSlketLr7CgauHvUdgNiTUCKFWMw81e/xpHtxGGDhwIJRKJXbs2KEtS01Nxa5duzBmzBhrnRmCsA9XNjCGwI9vAicXSGsT+lLNXq0D0PZ9aW0W1gfWvg4s72raODUUvJBetzAf+LoM8E050kIVd/56E/i+JvDiub1HIh+n4m+RUvy/gbUoyAHmBNvn2J8lAm6ljFbz9PTE0KFD8eeff2LgwIEAgPXr16NSpUro3LmzlQdJEHbg6DwgN920tt2/AU4v5t+nVgOacAfZT5j/n9wy7TgAEH8R+KMr0OFjoOsXxuuzhbTUaCCgvunHJuyLJtTA7V1A2HD7joWNSgkc/wGo1Fy4TkIkUK2j7cZkBUhTU8yZMGECDhw4gISEBADAn3/+idGjR1M8mpKOWg1c3Qw8i7H3SMzj+SPutloJKPP46/Kidx80HsxfzZLakXO/MQINwMTOMYZaDRydq9te1gaIPWW58RD24dJa6x/j6X3gyR1pda9uBo7OYRK7CuFWvN25ARM0NcePH8f333+PyMhIJCUlYdu2bXjttde0+7du3YrffvsNkZGRePr0KS5fvowmTZpo9z979gxfffUVDhw4gEePHqF8+fJ47bXX8M0338DX11fwuDNnzsSsWbM4ZQEBAUhO5klaZwlcvRiNiT1wlZ6mISwsDKGhoVizZg169OiBa9euYefOnVYcHFEsiD4AbJ3AfJ5pombDnqjVwMNTjCqfTc5Tef3U68Pdfu1XwLcS140VAApfAC4WiCaenw3snSqvzb3DhmWregMzn5s/HsJ+PDoL/DebuSZ6zjVeXy7KAmBJU+azFO3+/s+M9+nibv647IxsoSY7OxuhoaEYM2YMBgwwjO+QnZ2Ndu3aYeDAgZgwYYLB/sTERCQmJuKHH35A/fr18fDhQ7z99ttITEzEP//8I3rsBg0a4NChQ9ptq+YuUigkLQEVBcaPH4+FCxciISEB4eHhCAmhfDYlnjusaLszfYufYHNjK/DPWPntgsOAxMu67Tq9uPudnICuXzJ/M1kvUYUvtT8qPZfWglx5xpN5mYZlhfniAtMWvu8pza6OKOJo8o+1eYcRptnkZgCRfwL1+wFlqsrr99Ia4PJ63faLNODhGSbu0tDNQPI1wDsQcHYHto4H2k0GXjwz3m/8RaBp8XYykS3UREREICIiQnD/iBEjAACxsbG8+xs2bIgtW7Zot2vUqIFvv/0Ww4cPR2FhIVxchIfk4uKCwMBAuUN2eIYNG4aPP/4Yy5cvx5o1a+w9HKIokHyNu/3iOeDpZ4+RmMaN7dLqla8NpN7Vbbf7gBFkYk8ytgFiy7CfpwCzKzCfC1+6iz++wa1z+Gug5xzJw9b2wyb2OPNWXUfguWmqjRBR9BDyXuNb3twyjtGoHvwSCAoFun8LxJ0FGrwOlK8pfpwd73G3C/MYgQYANgwyrH9qkdGhAwAurQZeFbA7KyYUCZua9PR0+Pj4iAo0ABAdHY3g4GBUq1YNQ4YMwYMHD2w0wqKNj48PBgwYgNKlS3OWAokSTMJF7vY5vSzWhXnA+eXMmnxRpFCi3UxQqO7ze5eYCcHFHajZFXB2FW/r4g54lmU+52Ux/+tHKj77s7RxaCjgEWrWDQD+GlJ0zzVhGoV5jGavMB/4913gwgrhSNd8wnX0Ad3npCvA6j7AkdnA0mbyx3LLgiYHp5cy4QWKKXYXap4+fYpvvvkGb731lmi9Vq1aYc2aNdi/fz+WL1+O5ORktG3bFk+fCq+x5+XlISMjg/PnqCQlJWHYsGFwdy/+a6LFivxs4M9ewMlF9h6JOGxD1LwsRgOx52PdmrypPLkD3NjG2MDcP8K8dUoxuhUKW7BtIvBbJyAt1ngfEd8D/X5mNDMR3wPlasgaOgCdSj7uNPO//vKTPk/uAPtnANmp/PsLRdy4sx7LHx9RNMl6wmj5vi7LaEgurwV2T2HsaPhQWHCqTXtoWHboK8v1f2AGcHUjo3W6vQfIlmnHZmfsKtRkZGSgd+/eqF+/Pr76SvxHiYiIwIABA9CoUSOEh4dj9+7dAIDVq1cLtpk7dy58fX21f45oa/Ls2TNs3LgR//33H9555x17D6fkcWUjY9BqyYeKtZjpCzyPY+KwnFlqfn8FucDPLYG/RwORq4C1rwGnfmI+s0m6AhyZqxNk9n4KzPIDMl9O8tlPgX2fMYLClQ1AUhSQKsGjo9X/GG3Lm38xn81h90fM/8ZySv3WiTl3GvX/kbnMeb1/hNnm09RoYAtyykIgIwl4JJKigW0zQZhGYZ74b2IqbKPbmOO6z2tf56+vb+BuTBNyaydweR3/vlV9+MtNpTbPsuj2iUzMpI1vAj/WZsqOzQfuHTKsW8SwW5yazMxM9OzZE6VLl8a2bdvg6mpEVaxHqVKl0KhRI0RHRwvWmT59OqZMmaLdzsjIcDjBpmnTpkhLS8N3332HOnXq2Hs4JQ9VMYsAu6iRYVl+DuAm0eNOWcgIHUGh3Af14a91n5/pLQv/9jLuRfoj4LVfdEth/74DDP8H+L46s81e6qneGXhwVNqYLAlfwDy2oa9GE/PoPBMz59g8Znvta4wxtpimhr3vm3LGx/LvJCBsmKRhEzyoVMCPdZhr9tMY48uRUsnLBK5tltfm985cY/2b28Trb3oZ36ZaR8CvMndfepy8YwtROwJoMwmo0o7ROAmhKgR+aQOk3GS2pycAT6OZl5GaXcXt1uyAXTQ1GRkZ6N69O9zc3LBjxw54eMgPzZyXl4dbt24hKChIsI67uzt8fHw4f45GbGws0tPT8fHHH9t7KCWPvCz57ru2QG54/znC95ABh75iYrDsm849DsezQuAhF7UeyGHVu3cQSLnNX/fhGfFxNB4iabiyKcgxLDv2nWFZTip3SU/bXkQrkP+y72PfmzY2Qh55GYxXUH6m5Zb+Ht80NNKVS/Qh6Z59QsuclmDQGkZocnIGJl8Xr6sRaABgbkVGSFs/APgpVLCJvZAt1GRlZSEqKgpRUVEAgJiYGERFRSEujpEenz17hqioKNy8yZyEO3fuICoqShtPJjMzE927d0d2djZWrFiBjIwMJCcnIzk5GUqlbj27a9euWLpUpyL/+OOPcezYMcTExODcuXN44403kJGRgVGjRpn85QnCZHIzGO8FNkUlvP3CBvLbSF031yxbXVgOXN3EXyczSbj9oZnc7V9a8dczFmCvw0fi+03hwTF+7yUpAfQ0iAmUF/5gPJ2OzObf78ETpyvXce0ArY4x+6jU6JfhDoTjo3GIPsQERrxhRMsixPpBQEaizktJCgmRus9P7wPJRoQPObDDDPiZuILxnMe+x87IFmouXryIsLAwhIWFAQCmTJmCsLAwfPnllwCAHTt2ICwsDL179wYADBkyBGFhYfj1V0blHBkZiXPnzuHatWuoWbMmgoKCtH+PHumih96/fx+pqTopNT4+Hm+++Sbq1KmD/v37w83NDWfPnkWVKlVM//Y8qCXmXSquOPr3sxnrBgB393HL+N7ybY1KJS/Zo4aky8br6CM02XsHMRPF0Xnc8nI1GZdRuQQ0BN6N5Jb515bfjzHWvGreb6gsFH+LjzkGzKssvL/L54Zlez9l/r+4Eri+xXC/BrUaiPoLSDEjtYMjkHQFWNoCuLmDKxgvbGAovCxlpQsQsm1KjQaOfscIo6Zcu2yi9wML6slrs+djxmYNYIz6f21n3hhKALJtajp37iw6MY4ePRqjR482ub0G/Tg3GzdulDpEk9AE8svPz4enp6dVj2VPcnKYh7ZcGyaCRfJ1IJ7nIXhrp31zvRTkMm+SpnByEVAz3DLj0NjGHJ0LdPpUV/70nmn9TfiPMQju/aPOoNdayI1YzOb+YXGbGmPU7wfs/YRb9jwOSI9nMpQDjH3PjW3AGysAd29dvQOf67RoxS3QoiX5ZyxznW0eATQbbbifneOLzaGZwJg9huUawefoHKBWD/njaTdZeowYIW5sA17/nX9fx6nA8fny+ov4HihX3bwxsRE6p3aCElq+xMXFBV5eXnjy5AlcXV3h5GR3b3eLolarkZOTg5SUFPj5+Vk3GrOjcvcAYxdyczv//n/fsa9Q889YQyPdZqMNvZH4iD1hjRExEVPNRRO6vdkY5s25Ugvz+9TQ+TNmwtLwn8DSEMDYU4jBF/RMDt4BwDvnAWc3YHETpuzhSZ22BgC2v838v+N9YCDr3FrCm80RYAvOfNf98R+ATp8Ylj88xbjqt32PicTLB9/SpBBj9wOVWzM2Y+YKNQB/NGAXT+CVGUDLCcAPtaT3Za6noD7nfgUaDQQenQNq92RsdOwICTUvUSgUCAoKQkxMDB4+LHrrhJbCz8+PojKbyoaBxuskRDKq7/CZtn97ubPbsKx0gO5ztU7MEogQuz8CeswFTv/EeHh0mibdK0oIjYbBVDqyDLGdnIEIHqNdcyhVnr88bDjXpTYxSr47a4ePDHNMGcOfx4Px9i7DshtbuUINm/zsYpPixeYcmc0INad5hMAzS5m/menM9X//P+5+Y+7+bDRaNA8LOaes5NESTXppTF+6AvBlGvNSdWWD6cfo+R2w71Pj9fTZN405nxnxQI85TEoIO0JCDQs3NzfUqlUL+fkyLt5ihKurK2lorM3yV5j/s1OB12RGoxUiM5kxevStyGyf+535XLe38batJzGu0Q0HMG9T86sJ173wBxNhV6POLsgFeslUbVuaJkOt2z97CYeNfnbz3zvJ63fENuDsr8braWgxXl7/AKNVeoXHDufqJqD5WCbDuYsHUNpfft+OzN5PDSNs6zO3kmFZnIhHnmsp4P3LupgumlAP3jI8C8XgC0bpzhKYnJyA15cxQf6iBOLbAIB/XeF9rd9m3Mc3vsm///XfgG0CQXIz4pn/93/GaItf+RzwLCN8LCtCQo0eTk5OJrmYEwSHqHVAlbZMvh8vkRgQxlApmVgbAPBZEhPrRWN38XmKbmnmCU+wuleXMPmexr0Mx14oQViPYgV8O/8bV6hZb+byiinITfQnl3p9+csfngImnjHdRsnZXXpcFPbvqMGrPOM2Lsbx7/mFml0fApGrmXhCQMmxsZHqBGFMoLm4Uv6xp94HXD0Zo/asxzrhwZi2tvePjEAr1QOLjSuP7WfnTw2Fmhpdgbq9gNhTQBcjmbprdTdtH5sLf3A1rDbGsQxHCMJamOI19u8k4QijUmGv42enMMsgGp7rvAURxxOevfFg7rZYpmgNYoay0fv5y5tYIEBcn4X85dZewnP1FJ70A+qb3q9PMNB6onid15YBE44YCjQA8LZEGyeVytCOCtAJNEDRCTVgTdRqJkq1JZCzZDr+P2B6vE7AeOs48OFN7m/a5l3h9mFmZMR24Xn55tOOeJZhBKeBfwLljdjeOLsAbnray8nXgWlx8l7O+K5pG0FCDUFIQY6RIBv25GIK7HX8iyuBbf/jbmvU0nyRcJ15hJh+PwONRDQu7O8p1dsj/qLxOmIENwXqSFhKsyWvLhHfX1lEg9PmXaBsNSZSa7+fgaF/89er2Q2oKJB7i20LJUZGPLA4TLyOMt80obw4Yc0gdUK0mABUasZdwnRyNnx5CBvB337SOV3dbt/IPz6fM4u7N/C/o8BbJxibOM+yQNcv5PX7vyPcbb8QXQwl72BpfZBQQxBFnLxM2x7v1i5gRQ8gleXNceonbp2zPwM/NWE+63tHNB3Jr+UIGw4MWA70+gEYsALos0h4DFJtMSqIrNPz8Y6eO3y1jswymT51esnr1xwGreFu132ZX4dPAOz0KTB2HzBoLX9f7SYz/ysUzPmu2p6/ntj5lepBwpf2Qp9DM5l6mlxbjojKDtqohv2l1atQF5ikp0ktHci9b9q9D3zxlMk0L4XXfxPeFxwGBDUGukwHPrkvfwm3fC1mSTT0TaD/H9x9w/9hlqH+d1S8D2cSagiiaCMm1PBFgjWXTcOYjL8rjMWOefkGHqAXRdiY23PLCUCjN4B6rwrXkZpZuNvXwLAtwKidxus2Gmjo4dNpKvfNrkZXoPcCRuiyFUFNuNsa1T6fxkYzSVTrCEDBfXstX8dQWOFbJuDTolmL878ztlhyIiMXN+yRg61KW+l1y+kt+1TrYFjH2YXJND9kA9DaiAdRwzekHdfU0CQu7sDrvwKN9Tw+AxoAw/5mBKdmYyx/XAtAQo012TIB+Gec46t+iwIZScD2SUCiCZFxjZGeAKwWmfzHH7bcsVQq04wGnfRs/p0kGqmKuWwrRLQFZVnBu3wqAbXCX07yIngHMa7uABD60qspYr6h+3G3WUCLcea7k8uhTBVuoECNIOLqodPaaNB4nXj6AZ8lApOv6fa157HHcHICPr4HTGFF+7WGIGwMU5dQiyIqJZB8jblfcp4Zpt8oaji7cD2hxF4Y6vYGes7h39f2feCr50x/9qaRhBAXdoCEGmuRm85kcr3+D+OSS1geZSET3VetBrZPZDx3fu9s+eP8NUTnssjm/cuMAV35WuIJFvfKiP3wtYlukPoxNMrVkNaOz4NCy0th/NkDQ0GLLQCwl0o0OZn07QjePslM6r4vXWVfXcLYFLRk2Qi9GwmM2A4ESlhSsQYVm+k+s980h6zn1qvNsjVy82ImmAErGDdqoQd9aX/GeFhDuZrGxxMsYG9jKhkiOblMQaVkYhtdFbAZsiZbxgO/tmeCES5qbJhCovUk5v/+f4gv1dTvZ9rxxYx/hfjgqu6zlDhC4/8zLAufVXSi91oyCKYFIaHGWrCTqRWFnECOyI73mFwop36ybs6b5Kv85WWr6964+4s8OI25kFoCjYeLwgno9wsQ0lJ6W43WRJ+8LCZEO58hKlszxH7IdpnBaK70vZkCG3HrObswNgXssvI1gRpdpI/b0kgNWMfnqt3oDeY7S36DljAxPWYlL7REVvJ7B83vg83N7Yz77tbxTMBJW/HgGBN8EGBi8uTzLA13/5Z54Wg8EAgVOXc95hrajRhjZjrQ41t5bQDGKLjHXKBCfaDzdOP1KzXjbofPsuuyjgEubpbxfLQwRegMORiFrGRqh7+23zgcGU30zENfcSfH+/8Ba17jD1hlTaYnMHlVptxigq/JRWmGXYBGU1MzHAiT+aDp8S3QfJxh+fV/dMn09KnbG6jS3jBbtpMzUKm59BgtRYkq7RgNUw8B1b8lkfK2PXQzq34RfFRnpeg+bx4BJEg0cjWXNSJLwQDgV4WZ/NlLfG3eZYxz2bj7MkEsGw8EPrxh+XHy0WYSEwm4dAVp9X0q6j63n2yVIZlFv58ZGzg2I/+1z1heUgQW5hwUdoZYoVxBhOXIZKnWNbFh/n0XGM0KMZ8QyWgYgkJldq6AdikGYOI28EWidS+ty6uib0dVmG88ToycMOxCbU0xQPUqC/RZAFyUYZjr6gmM4UnLUJxRKIB+NsqhJMXOrhIri7Sc8Pdhw4GKzYFdk2UPSxKaset/h6QrTCqJUv6MfZVazWjf9EmNZpYhRZc+zeAtnlQgPb4Fus/mxrIJZj0HNMuifJiSNNJSjNjO/I6d7BfMThSFglmeTbgEhLRi7LbcS9t1SCTUWAsp0VsJ68JO0nj/CLD2NebzF6mMJuH5IyavTo2uwN6pjH1HXT43Yr2Ht1+I8WPrh0ePOwNUNxJqX6pQ8+Ym4C9WYL20WJ1m0BZeNR/dlVbPqxwTzE8oTkdxIrCx8DKktdA31A4dytjqvf4rME/gGnzrhG6pT60Cdk+x7JjUamBVH0Yjpx9h9tBXzPjYzEjmCi8PjgJr+jFjfPuk9OO+eM4EmKzZVbxet6+Fw/Pra8eMeddV7cA8E2r3BKI2MHZ1XXgiOFsT/9r82cOLEq6eQNV2zGdn+wo0AAk11uPBUe52EUvPXmJ4dIGJ2aARaABmeapqe2BRQ27dB0cYb5xWrPwmpnqu6a99i3menFzIBLBjL1mKEdyEu/0T642Tz33Y0nhLDAz39inmzb2IeknIwtJReaXE9mH/lr6Vmdw+xghqrPvcsD9XqPGrzHyPwlzhnFdCqNWMwfu9g7oIxiGtuHX0BRoA+DYQaDJclwct6i/m/+RrzFikLlNuGMyEOOgyQ7yesfgo/vWAJ7cYgVF/CajvYmDn+8zn4DBGA6FZwhr5L3B1I+OVRxRpSKixBs8e6PLzaMjPkv8gIcxnRTjz9spmwyDmjY6PvVOZ+BMaDxy5mZnZTIsD5lVmPnuKhBiX644qFm02+4m8vqyJTxDQ1AG0NIDlgruNPcB4Rb4iIcqrkxOTSuH4D4bB2wasALawJli+VA/6WrvnccDPLZnn09QYZtlR49X2yX3+jOUZicCCevzjk6pZjFqnE2rUKl35nb1AfSP2MRoevfz+R4wY6BqzPZp0hlkmq8CTAqPZKCZoZfYTQ4GnfE3+PFtEkaMIWp85AKt53ARfpNl+HATDno8Ny8Siq6Y9ZP4vyAXW6wW5ekNGsjsPX6DMy6zYqgJm3fnY98DujxkNkhifPhTeJ6bxM8fDRfMWzM7+SzBITVtgjMqtmCSGfBGU+WgyFHj/kqE9VqM3GKN0MVx54vxotCz3/wOSWMtp6wW0aUICDQCcWiR+fDaPbwJ/9mYEOg2bR+g0oYX5zF92KvDXUODvMcDpJcDdA9KPARiPlaRQMJpOIfs2hUK6ES9RJCFNjaU5uRBIjzMsz3nGqH8J2/PonGHZ2Z+F65//DajXB4g/b7ivdk95x9YYzf0ZwS2/sFw8e7L+pNfvZ+Dfd3SB3novsLy9RKepQIePmbdi/fECTEZrPi+pkkC/n4GdHwDtPrD3SHQ0H8MIy9UEbLUUCiaWSDyPAK1Wc8uzrJxCYcMgJqqxPgmRwLHvgGge4UXjtl1KYrqOoFD5KTsIh4OEGksjtJSgn5uHkM/T+8D55YxNR8u3jNc3lZjjwNXNwJM73PK3T0qPZaLBUjmjwoYzfxpajOMXakqZ+Zbp5GQYnVhDxPfMklJJpGw1YJQNY7FIwdkVaGMknL5PRQACWkG2zQ47QjSbZqOByFUmDE4PPoEGYNzC+QQaNlKWVD+4ykSFJko8JNTYirPLgBqv2HsUxZslrAir1g6LvnWCYZkpkW7FYuUoC43H3BCj4QDDSKoD/zS9P2MUx9gzJZ2e85gM7tH7ueUpN4GTC3TbsScY4aXZaF1Zfo7loxDrk2ih2DbWcg8nih1kU2NJxB4A0QeA499THigpFOYBD0+b53FiCy8gc0m+Cjw8ZXp7vvxBQhmhLQEJNcUPnyBg2GbDcrZAo2HnB4w92YkFQH428H0NQ2FICr15+hbiuBG7IKmQEwbxEhJqLMmW8dxtf7313f9m869vE1xmV2BsOvi0JVJo/Q6T+K2oU/BCfP8bK5mop+8IXDNSk1bKpaxA3ihbZpYm7MNPjYHDs4A5waand2k2Bhht49gqpKkhXiJbqDl+/Dj69u2L4OBgKBQKbN++nbN/69at6NGjB8qXLw+FQoGoqCiDPvLy8vDee++hfPnyKFWqFF599VXEx/MkDNTjl19+QbVq1eDh4YFmzZrhxIkTRtvYlId6waTe5nkLL0out0URtibrxjZm+8Y2xp7GGG+dADpNA8K/st745FK+jvC+VXyB/gAEvFzmajgAmB7HBOCyJaXKMckl348CPnmgKy8O2i+CH0sJpGLJITU4OdnmWvn8CfBZkrjBPVHikC3UZGdnIzQ0FEuX8ocTz87ORrt27TBv3jzBPiZPnoxt27Zh48aNOHnyJLKystCnTx8olUrBNps2bcLkyZMxY8YMXL58GR06dEBERATi4ng8jYoKvMntKACfIImXge+qcsvOLGXyD7HtafT56A7w5TMm8FiX6YCLu2m5crqbkKTOGBHC9wEvZasDE3iy8/Khb09jScrXZIxjNQLO5OsUPLI4Yyx6rhS8g4HGg43XA5hrRwyxuE1SeONPxi3bjcdtnSjRyH7yR0REYPbs2ejfvz/v/hEjRuDLL79EeHg47/709HSsWLECP/74I8LDwxEWFoZ169bh2rVrOHRIONDZggULMG7cOIwfPx716tXDokWLEBISgmXLJETZtAXHBNaG9V1gSVPDT2EesHE4kPucW35AQsAr70AmbDubljKXrlw8gDo8bswa3HnsV6RQRaaNy6hdxnNEachJNWxrDcrXlJYagii61OtrXvuP7jJJH6UKtl5lgYmnhffre9iNPyx9LHV6MdGSCYIHm9vUREZGoqCgAN276/KGBAcHo2HDhjh9mv8myM/PR2RkJKcNAHTv3l2wDcAsc2VkZHD+rMaR2dztNzcx/0foJULbWQxsPWxNYT7wUxMmt4ql4IuOCgDVO+s+e5UHev0AVGjA5FdxEQix7lMJGLvPtHHINa6V4zLOXlKYmQ5U6yDvWETJwVwtWyl/w9QfADB2P/OMe+s4kytpJMvtPaCBcH/ZKdztgIbAx9HA4HW6gJV8hA0H+v4kb+xEicLmQk1ycjLc3NxQpgw36VhAQACSk5N526SmpkKpVCIgIEByGwCYO3cufH19tX8hIVZ621zJE5CtwstInHxLUMpC64yjuPLsAZCZaP3j1OzGfUP83xFGozPpNFCxmXDemIkngQCesOpSkDuZSI00C0gLtU8QGpoM5y+v2U28Xdev+AUaAKjcmsmVFhQKjN5lPGmrEC7uTCTfen2B9y8D5Wrx1+v3M0X8JUQpMt5ParUaCiMTgP5+Y22mT5+O9PR07d+jRwIBoMwl7gx327OsePTgA0aSspU0hIK9mYv+g9G9NDfni/5vJLTsI5T1Vypv/AmEzwQaDzGvH33I44OQw6tLdBGp2RjLfN1uMnd7yF/MM27YPxYbGkf4VyiA7t8w9+qE/4AaL8f30R3+tgTBwubB9wIDA5Gfn4+0tDSOtiYlJQVt27blbVO+fHk4OzsbaGVSUlIMtDds3N3d4e5uJGurNRh/SPwN/dyvQMR3thuPPZCTlbwg2zpjGLWDm7smbDgQ0hpQFQL1efJzGcvwayqa9f/ncUymXz5cSwEf3ZLXb8Vm5o2LKFk4OQHeehGhm44CWkwA9k0Tb8embi+gzgPLGY7zBbWsE6GzcRux1TLHIUoENtfUNGvWDK6urjh4UJd4LykpCdevXxcUatzc3NCsWTNOGwA4ePCgYBu7ou/OGNjYsM4CE5czigOzA4FZfky+K2MU5gO/GUlCZyo+wUDdPrrt4KaMtqbnXEZtrg+fTU1rI2Ho5eBWWnhflTb8wfTEqNgUGLUT+OCKeeMiSg76WtFXFwt4ab4koCF/uVSBZsgG43UmHJHWF0FIQLZQk5WVhaioKG38mZiYGERFRWldq589e4aoqCjcvHkTAHDnzh1ERUVptSy+vr4YN24cPvroIxw+fBiXL1/G8OHD0ahRI47HVNeuXTlu41OmTMEff/yBlStX4tatW/jwww8RFxeHt99+2+QvbzX0hZqxPFE5MxKA9ATbjMeWqNVA4cugcvONuHUCwJW/zDteVyMxadjGtF5G3EgVCsZQUcOU20APC7p5iwk1DUz05qjWEShT1bS2RMlDSBgRCoHQ/kPzjle3N3/5+MPM8u/AVRSpmrAospefLl68iC5dumi3p0xhkuqNGjUKq1atwo4dOzBmzBjt/iFDGDuCr776CjNnzgQALFy4EC4uLhg0aBBevHiBrl27YtWqVXB21rnl3r9/H6mpOpfVwYMH4+nTp/j666+RlJSEhg0bYs+ePahSpQgmMXPXm7zcvJgorc/0Ashd3wK0czBvKJVMI2ihRHpS6WAkU3VQqC7brxTq9QVmPAbUSvnJK43h4gZ0+RzIywBOL9aVv/IFEPqmZY9FEMao/5ruc+8fgV08AkxIS8sec+QORhBXKPjtewjCTBRqdclJRpSRkQFfX1+kp6fDx8fHch3P1Fs24ItwubipoVDT7RvHE2ruHwHWvqbbNhbtM/aUcGRdIQavBzYNA8JGAP34g0BqKcwHTvwA1OoOVGou7zjWhH3NUERUwpY8ewBkPwVCWujK1GomJtSZpUxm74wERgNoiaXNk4uAQ18xS8FD1pvfH1EikTp/U5ZuS9B9ti5IXLMx/HV41bsOKE/+pefho1IJu4MCQGEud3v4VuDoPCD+PLe83y/AvUNAk6FArW7SBQEXN6DLZ9LqEkRJoGx1Qw2pQsEstWqWW1+kAe4WevFrPxloOtJ8L0KCkECRceku1rR9j5lkP7oD9FnIX0fNkwJCJZwWoljy7IGhkLKwPpCfA8SeBJa/wqRCYHNS73zV7AqM5xqEY+x+IGwYMPBPRqBxBGryR9wmiCKBZxnDKN3m4FWW0mwQNoGEGkviHSh84/pWMixT5jP/v3gORK5m3o6KM4vDDMsyk4Crm4BVvYGESGD9IO7+WFZSUr4EoADg4WexIRYZ7skIC08QBEFIgoQaW9HvF8Oywjzm/3/fYdInfFcVWBkBpD206dDM5sld4MEx4f27Jus+Z6cAf3RjvvuZn7n1AgXcRx3Ru6dcDXuPgCAIwuEgocZW+IUAX6QCnVn2HZrklrdZiQjjTgM7P7Dt2MwhPhL4uQWw5lUZbc4z33m/RFsXVw/jdYobb25iAgEO/dveIyEIgnAYSKixJc6uQOdPdduX1wJP7xvWK06ZvG/tMF6Hj+yn4vtf/820fosL5WsC4/YDtbsbr0sQBEFIgryf7M2SpoZlapV4m4IXQGYyUFZCcDtrkZsOxJyAyR5cCRfF9zcaxLiZWjpOBkEQBOGwkFBTFCnIEd//W0cg9S4w/j+gkp3y/6zqAyRfNb391U3c7fH/cbednIAmFJCOIAiCkA4tP9kD/3ri+9Ni+cuVL6P1pt5l/j9kJEWAtcjNEBdoRu+W36e9hDOCIAjCYSChxh60lpCvKlbPvfniSuCbcsAKlg0G2x3aFiRdYSLhHjOSYbxqe9uMhyAIgiBYkFBjD8JGGq+jnzpAk5fl0TnLj0cqmmzaZ4ykJgCAbl9bdywEQRAEoQcJNfZALG0Am5iXmpjMZOuNxVq0+wCY7oBZyAmCIIgiCwk1RRmNa/f+GfYdh6m4enK3m44CJvFomqbcss14CIIgCIeGvJ/sxeu/Adve0m17+DHZcVNu6MpynzP/l/IX7icvC3AvbY0RcnkWI7+NkzOTUbsgB6jbG3ArpTN2ZuMTbP74CIIgiBIPaWrsRS29oGvTHgLjD3HLdn0IPLkDnFsm3M/8akDkKiDnmeljyUgCbu0ST7CZJlGoeeNP7na9PkDjQYxAAwDOLkCHj00bJ0EQBEGIQEKNvXBhhf4PeJnzSH+5BgB+NhJ8TpnPpFXY/NL4+NZO4OdWQPJ16WNZUBfYNAzYN024ztrXxfvo9zPwbiTQsL/04xIEQRCEBSGhxl6whZo27zD/KxRAtY6m9Rd7Atg4DNg0HHhyG9gyTn4f53837dgAEDacCf0vhev/mH4cgiAIghCAhBp7wfaAcvXSfR7xr+l9shNjPrktrQ1f7ik5VO0ANBsjr41viHnHJAiCIAgeSKgpCrCFGicn3XKUuSRcAh4cFa+z9X+m9990JDB6F9B3kbx2wygzNUEQBGF5SKixJ698AdTvB9Tsyi2v0cV42y4S3LyXdwHW9AMyEg33Pb4J/NLWMLHkg2OGdfOzDcs+uAL0+cn4GPhw9QRajGc+O7ma1gdBEARB6EFCjT3p+DEwaA3j+symuQR7mAsrpB/nGo8Ny5bxXPdxDWteNSxLjNJ9Hrwe+Oo5UKaq9CCCfHT9CujyOfCOHSMkEwRBEA4FCTVFEbYRsT6TzgIzHgNZMqIMH/wCyE1nPt87DFz4A0iPl97ew1f3uU4vxqDZXDx8gE6fAOVqmN8XQRAEQcAEoeb48ePo27cvgoODoVAosH37ds5+tVqNmTNnIjg4GJ6enujcuTNu3NBpBGJjY6FQKHj//v5b2NZi5syZBvUDAwPlDr94UDpAeF+FeoCriNAjxKYRQPxFYF1/YPdHQF66cF21Wvf5yV0mDg4AlKpgnnaGIAiCIKyI7BkqOzsboaGhWLqUP6nh/PnzsWDBAixduhQXLlxAYGAgunXrhszMTABASEgIkpKSOH+zZs1CqVKlEBERIXrsBg0acNpdu3ZN7vCLB05OQO8FhuWjduo+Nxoor8+YY8CVv6TVvfY3oCxgMoX/3AK4sJwpd/MSb0cQBEEQdkR2moSIiAhB4UOtVmPRokWYMWMG+vdngrCtXr0aAQEB2LBhA9566y04OzsbaFi2bduGwYMHo3Rp8XD/Li4ujqud0afZaGD3FN32p7GAZxnd9mvLGOFDDlmPhfd1mQEc+Zb5fH45kHwNOL2YWyctVt7xCIIgCMKGWHQtISYmBsnJyejeXZcCwN3dHZ06dcLp06d520RGRiIqKgrjxhk3jo2OjkZwcDCqVauGIUOG4MGDB6L18/LykJGRwfkrNjg5M8s9GtgCDQA4m+A19OSutHrx5w0FGoIgCIIo4lhUqElOZoxXAwK4NiEBAQHaffqsWLEC9erVQ9u2bUX7btWqFdasWYP9+/dj+fLlSE5ORtu2bfH06VPBNnPnzoWvr6/2LySkmAV9G/Y3ENgIGL7VMv2l3uEvb/gGAAsY/xIEQRCEHbGK1adCzztGrVYblAHAixcvsGHDBklamoiICAwYMACNGjVCeHg4du/eDYBZ3hJi+vTpSE9P1/49evRI5jexM8FNgLdPGsax4aNuH9OPU/ACUOaZ3p4gCIIgigAWFWo09i76WpmUlBQD7Q0A/PPPP8jJycHIkSNlH6tUqVJo1KgRoqOjBeu4u7vDx8eH8+dQ9JjL/N/+Q2DIeuCLVOD134D3L8vsSA1UbmPx4REEQRCELbGoUFOtWjUEBgbi4MGD2rL8/HwcO3aMd3lpxYoVePXVV+Hv7y/7WHl5ebh16xaCgoLMGnOxps0kYPJ1JpAdwNjZhA4BylYHPksC3MQNr7Wo1UwOJ2PU7mn6WAmCIAjCysgWarKyshAVFYWoqCgAjHFwVFQU4uLioFAoMHnyZMyZMwfbtm3D9evXMXr0aHh5eWHo0KGcfu7du4fjx49j/PjxvMfp2rUrx238448/xrFjxxATE4Nz587hjTfeQEZGBkaNGiX3KzgWfiH8wfDcvIA3/pTWh7s34OJmvN4rn8sbG0EQBEHYENku3RcvXkSXLrrcRFOmMG7Ho0aNwqpVqzB16lS8ePECkyZNQlpaGlq1aoUDBw7A29ub08/KlStRsWJFjqcUm/v37yM1NVW7HR8fjzfffBOpqanw9/dH69atcfbsWVSpUkXuVyg5KCTIrOVqAuEzmc9VOwCxJ/jrvRsJlK9psaERBEEQhKVRqNXs8LGOTUZGBnx9fZGenu549jV8PL4BLOPxKitfG0h96d49kxVZOOU28Esr/r5mikQgJgiCIAgrInX+lq2pIYoRFeoDrqWAAr0s2/87Ctw/YpgNvEJdmw2NIAiCICwNJfJxZBQKYEYitywoFHArBdTrw/yvT8ephmXmuIsTBEEQhI0goaYk4MRSyPEJLWy6fMbdHrMPGPCH5cdEEARBEBaGlp9KAjMeAw9PAlkpQN3e4nX1PamqUPwagiAIonhAQk1JwNkFqN7Z3qMgCIIgCKtCy0+EIdU62nsEBEEQBCEbEmoIQ9xLgLs7QRAE4XDQ8hNhSM+5QFos0HqivUdCEARBEJIhoYYwxK8yMPGUvUdBEARBELKg5SeCIAiCIBwCEmoIgiAIgnAIStTykybNVUZGhp1HQhAEQRCEVDTztrF0lSVKqMnMzAQAhISE2HkkBEEQBEHIJTMzE76+voL7S1SWbpVKhcTERHh7e0OhHznXDDIyMhASEoJHjx6VjOzfdoLOs+2gc20b6DzbBjrPtsGa51mtViMzMxPBwcFwchK2nClRmhonJydUqlTJav37+PjQDWMD6DzbDjrXtoHOs22g82wbrHWexTQ0GshQmCAIgiAIh4CEGoIgCIIgHAISaiyAu7s7vvrqK7i7u9t7KA4NnWfbQefaNtB5tg10nm1DUTjPJcpQmCAIgiAIx4U0NQRBEARBOAQk1BAEQRAE4RCQUEMQBEEQhENAQg1BEARBEA4BCTUEQRAEQTgEJNQQBEEQBOEQkFBDEARBEIRDQEINQRAEQRAOAQk1BEEQBEE4BCTUEARBEAThEJBQQxAEQRCEQ0BCDUEQBEEQDgEJNQRBEARBOAQk1BAEQRAE4RC42HsAtkSlUiExMRHe3t5QKBT2Hg5BEARBEBJQq9XIzMxEcHAwnJyE9TElSqhJTExESEiIvYdBEARBEIQJPHr0CJUqVRLcX6KEGm9vbwDMSfHx8bHzaAiCIAiCkEJGRgZCQkK087gQJUqo0Sw5+fj4kFBDEARBEMUMY6YjZChMEARBECWYyMeR2Bezz97DsAglSlNDEARBEARQoCqAq5MrAGD0vtEAgJp+NVGzTE07jsp8SFNDEITDEp8Zj1+ifkFabpq9h0IQRYZ/7v6Dpmub4uijo5zyxOxEu4zHkpBQQxCEwzJi7wgsu7IMM07OsPdQCKLIMOvMLADAh0c+tPNILA8JNSWUG6k38Pq/r+NUwil7D4UgrEbqi1QAwMXHF+08EoIogujZ3KrVavuMw4KQUFNCmXhoIu49v4e3D71t76EQDk6hqhDZBdl2HYMjPKwJwtIo9KUaB4CEmhJKZn6mvYdAFEFyCnKwLXobnr54arE+B+4ciNYbWlu0T4IgzEdfqFGj+Av/JNSUVBxPQCcswPwL8/Hl6S8x/sB4i/V57/k9AMDpxNMW69Pa5Cnz7D0EgrA5jqDRJKGGIAgth+MOA9AJIlY5xsPDWHp5aZF9gB5+eBjN1zXHupvr7D0UgrAqCoUCOQU59h6GRSGhphiSp8xDQlaCvYdBOCDOCmft523R26xyjMlHJ+O3q79h14NdVunfXD498SkA4LsL39l5JARhXfKUefjx4o/abVp+IkzGnLfUgTsHoueWnrj25JoFR0QYY1HkIsw5N8dsDUNGfoasPi49voRzSefMOqZUnBS6R8KXp7+0aN/64c1tKTQ4wsOaICyB/rPnVKLOA9YR7hMSamxIUlYSnuU+w/wL89FnWx9k5WeZ1E9MegwAYF+s6WGtHdHq3ZooVUqsuL4Cf93+C/ef3ze5nwvJF9Dur3baOBHGKFAVYNS+URh/YDwy8jNMPm5RJLcw195DIKxMel66geebUqW002gIwFBw4Qg5xV+mIaHGVqTnpaP7lu7otKkT1t5ci7jMOGyN3mpWnySY2A6VWqX9bE7UzZ+jfgYAbIneIql+gbJA+zkjr+gLNezzpI/+9apUF83JzZGNhHMKcpCSk2KzY7Xf2B6tN7TWlsVnxqPDxg746dJPNhnD1SdXEZUSZZNj8XHlyRVcSL5gt+PzoX+Psp9nRfWelAMJNVZkX+w+/BL1CwpUBVrtChtzVX1ZBaZpegj5sG/23678ZnI/cpeu2NeIsey09ub7C9+j86bOeJLzhHe/vlAjVSjffGczBu0cpA2kR5hOx00d0fXvrjYRbB5mPDQoW3ZlGTILMvHHtT+sfvx8ZT6G7RmGEXtH2CVOklKlxPA9wzF2/1ik56Xb/PhCiM07Yi8lxQUSaqxETkEOPjn2CZZdWYbx+8dbZUKS+rZPMPx48UcsvrTYoDw9Lx3fnv0WV59cFWzLvtkLVAWC9SwNWwgq6pq5NTfXIC0vDWturuHdr38PsO13xPjm7De49ewW728nlaLqaWVrNFqoyymXLdanWq3mvSfYv7c9zn+uUre8aY+4XOxzIvf4x+OP43j8cUsPiUHkp8gpLP6eUCTUWAn2BX0p5RKcitipLuoTJBtLPBCf5DzBqhursPzacgNbjoWRC7HxzkYM2zNMsD1bqDFHwyZXuGUfS6oQUFQx95oz54HrCAaQlsSS9//7/72PThs7iU7c9tAAsJ+59hCq2N9Zzn1foCzAO4ffwTuH38Hz3OcmHTsqJQr30vjDMqgg/Fv8fvV3k45XlCjeT8liRHGfkGyJSq3Cs9xnABgtSvg/4fj27LcG9XILc3Eh+QIKVYVG+8xX5XP6ZxObEWu0PXv5yZwJwZzlJ1tT2rW0ZTs0cx41Z2IkoYaLJTXHR+OPIrMgE//F/cc9BusHF5tIpXAm8Qw+PPKhyUuQ9vj92dernJfaQrXueZaWJz+7fEpOCkbsHYHXd7zOu1/sGeQIoUJkz7THjx9H3759ERwcDIVCge3bt3P2q9VqzJw5E8HBwfD09ETnzp1x48YNTp28vDy89957KF++PEqVKoVXX30V8fHxosedOXMmFAoF5y8wMFDu8O0G2b9I5/3/3kenTZ1wOeUy/r77N1JyUrDxzkaDep8c/wRj94/VGt+aiquTq9E6nGUgG9q22HP5qUfVHlbt39rf59LjS7oNG85pK6+vxMGHB213QABZ+VnYcGuDoD0TAESnRWs/W+Pciy3Lmqsp+d/B/+FQ3CHMOTdHchu2IGMPoYbzIiTjmcE+V6YI8sYEE0cX8GULNdnZ2QgNDcXSpUt598+fPx8LFizA0qVLceHCBQQGBqJbt27IzNSpJidPnoxt27Zh48aNOHnyJLKystCnTx8oleKW1w0aNEBSUpL279q14hOnpSgv99xIvYHeW3vj0MND9h4KAOBY/DEAwPpb60XrHX10VFI9QPyhKkWoYT+gbLmUaC9hCjBduyh0ratU3Ae03O8j9wF/LVX3fLDVg/xhxkMsjFyIKUenIKcgx2bLLrPPzcbc83Mx7sA4wTpsGw1rPI9sYYAanyn+8it0THssP0nRIBvDGteP2LkYXGewxY9na2Q/tSIiIjB79mz079/fYJ9arcaiRYswY8YM9O/fHw0bNsTq1auRk5ODDRs2AADS09OxYsUK/PjjjwgPD0dYWBjWrVuHa9eu4dAh8UnVxcUFgYGB2j9/f3+5w7cbRW35iT2hzDozC3GZcfjw6Id2HJEhCihw5ckVSfXkoP+gcHZyFqipw9ZvN2q1GicTTuJxzmNtmeZ7KlVK5CvzhZpaDHOuWb4YTPpv8nJ/N7kTE3v8fL+fSq3Cjac3LHou2W73rTa0woAdAyzWNwAcfHgQU49PNQhtf/wRI7DweVlq4AiRVpCP9X9P9vnX3HPmClOmCvbGrp2s/CzsuL9DViwodrgFPtgvQnKuXfa1ahWhRuRZ1qh8I8n9pL5ILZIG+BadaWNiYpCcnIzu3btry9zd3dGpUyecPs0ks4uMjERBQQGnTnBwMBo2bKitI0R0dDSCg4NRrVo1DBkyBA8ePLDk8C2K/sXIdyEVlQuiqFq8K6AwiM1yIfkChu0ZhltPb+nqKRRIykrCR0c/EoxJwXlQ6K3vs1MDCMEOGGaOgBOXGSep3t6YvZh4aCLe2PmGwb43d7+Jjps64kXhC5PHIQVTJ6Dt97ajzV9tsPrGak65uYKh3Pbs8fNp19bfWo8hu4ZYVJjXn3QtnUNrytEp2BuzF6tvcs+tk5Ph90vISuDYoLDPhzU0NfrLHuxjmPP8M3UJVo5x/4yTMzDj5AxMOTpFUt93nt1B03VNOSkGxI4vx6aI/X2tETdGTFCSeo/9F/cfumzugi9OfWGpYVkMiwo1ycnJAICAgABOeUBAgHZfcnIy3NzcUKZMGcE6fLRq1Qpr1qzB/v37sXz5ciQnJ6Nt27Z4+vSpYJu8vDxkZGRw/myF/sXxybFPDOrIeevIKcjBvPPzcDH5otljKy7EZBi+dY7dPxZXn1zF24fe5pRPOzENBx4ewIi9I4z2q/8wlaKRYD8IzDGglWroqFmCY6O5pm49u4XsgmxcT71u8jikYKqmRmPc+MPFHzjn+kG6eS8h5qjz+e61tTfXAoCo66yxt3Epx7EG+nFm9OOgZORnoOeWnuiyuYu2jBPzyApCzR/X/hCMkv4877lBWe9tvSVFlWZ/VznxZuTY1Pz3iDFylpqOZMnlJQCAVTdWCdZhX69yNC7mamqM/bZi7vxSBc1fon4BAPx7/1/pA7MRVlkT0b+x1Wq10ZvdWJ2IiAgMGDAAjRo1Qnh4OHbv3g0AWL16tWCbuXPnwtfXV/sXEhIi41uYh/7F+DRXWPiSworrK7D+1nqM2T/GrH74KCoaI31uP7uNUP9Q3n0a7yiAuYkfZT4S74z1FQ2WnyRoathvWueSrZ+Hie83UavVnPKx+8daPCkk523eAhM0+wH95/U/OeOXKzSdSDiB80nnJddnj5/vQZ+UnWS0j7W31hqtsy9mH47EHQFgO3urf+7+I7o/KUv33QqUBUjPSzfZxVgOnOVS1jF23N9hUPdR5iMciz+GxZcWo++2vhazQTmZcBIFqgLO97VHpFy2wGbq8pM1xi2WlVuKpib1RSrupN2x5JAsikXvQI03kr7GJSUlRau9CQwMRH5+PtLS0gTrSKFUqVJo1KgRoqOjBetMnz4d6enp2r9Hj4xMfBZEykUs50Lni85pDuyHvJSHu72o5ltN+1lI1SnlrVPs7UeSpkZlHYPP3MJc3uuAT12thtrgITf9xHTt5wfPH2DPgz0WE1It8Tavf67Z25V9KsvuT5NBWy6map3Yy5x8PH3xFJ8c/wTvH3kfSpWSV1iwx0sDW1Aff2A82m9sj7vP7mrLrOW4wP6uHI9PgVOggALLry1HbEasYHBFOc+nVza/gomHJqLp2qZcLyIz7l9Tg21uvrtZd3w5mhrWuE25djgaKp72YoKSlONNPDRR9phsiUWFmmrVqiEwMBAHD+rcGfPz83Hs2DG0bdsWANCsWTO4urpy6iQlJeH69evaOlLIy8vDrVu3EBQUJFjH3d0dPj4+nD9bUdTd5tixEGwZIdcctt/bzr9DYfwhzb6RTdHUiD0IHqQ/MCliaWJWIlqsb4GPjn1ksI/vIVigKhD9rfr92w+fnvhU6xVmCpYO9qd/H7CjvPK5jP9992/REPqmxikxVTNh7N5g/+5CExf7O7NJy01Dv+39rJIygP19L6Uwru17Y/fy7rckV1OvaifG2Wdna8s1QrrYcc8mneUtP5+s084Ze66yNeJidnRSiU6LRot1LbAochGnXIpQ+Nftv8w+vinLT4+zddoyPrs7c21qbj+7LbgvNj0WDzMeWkTrZiqyn1pZWVmIiopCVFQUAMY4OCoqCnFxcVAoFJg8eTLmzJmDbdu24fr16xg9ejS8vLwwdOhQAICvry/GjRuHjz76CIcPH8bly5cxfPhw7bKShq5du3Lcxj/++GMcO3YMMTExOHfuHN544w1kZGRg1KhRZp4C62Bpq3WxEP530+5i1fVVBuv/Bx8exIAdA3htGeRedPti9lk0tLq5eLt6az8rGKlGFLHYD5I0NQIPpdvPbqPf9n7o9k83o33os/L6SgDgjWnC98aUV5gnSQC9/tR0W5u0XJ0G1SJCjd730NixAPxLNV+f+Ro/XfrJ+HKijTgcd1hrP8BGrVZj9tnZnPhJQoKv0L228vpKPEh/ICm5o1KltGtenuup1/HDhR+M2rR8dfor7I/dD4A7+Ql5P0kRrjjLZi/bH3t0DL229hJNVskx1DXx3P106Sco1UqsuL7CpPYa5GhcpC6b5RTkoN/2fph3fh6nnL3U12trL+3nAmUBdt7fyVkiNBinmS/jr25/FX229eG1obIVLnIbXLx4EV266IzPpkxhrMVHjRqFVatWYerUqXjx4gUmTZqEtLQ0tGrVCgcOHIC3t24SWrhwIVxcXDBo0CC8ePECXbt2xapVq+DsrHtjvn//PlJTdW9l8fHxePPNN5Gamgp/f3+0bt0aZ8+eRZUqVUz64taiUFUIFycXs1TOBx8exPcXvsf/Gv8Pb9RmvF+EVLBqtVrrNlqoLkTroNbYcGsDPmj6gdaSf/qJ6djUZ5PJ43nw/AE+Oc4YOl8bVTRiA+knehR7c1KqlOj3bz+DtglZCdgfu19Scj8h9fXJhJMA+A0Yrz65ik13NmFy08nw9zIMP7DpjuFvkp6XjkuPL/EKL3mqPFnCqFKlxLnkc2hUvhG83byNNwBXg2eJt3n9yUQszgj7npHjWisE+1zpXx9yJrllV5ZhUpNJnLJLKZcMfj+lWsm/lChwLKka0tQXqei7rS+q+1bnlF97cg2N/PldcI1pEuQuP725+00ATO6oGa1niNbd/WA3ynmW45RpzsuZpDOC7YQmVL5z+u5/7wIAJh2ahNND+b1mLeFFJNiOdfrylflwc3YT7ccahsK7Y3bjQfoDPEh/gGktp2nL/dz9tJ/ZmqvVN1cbFaDNXSotCisUsoWazp07i35xhUKBmTNnYubMmYJ1PDw8sGTJEixZskSwTmxsLGd740bDiLJFjb0xezHtxDT82OlH1C9X36Q+Tiec1gojs87M0go1Qqy7tU77+ebTm9qLli2Nm5uhNj5LesAre+AEJ9EJ+G7aXc625kExaOcgg8nzbNJZbLq9CW/WfRMtg1rq2pigPtbkknqW+wzLwpdJajN632hhN2C1NG8czYS19uZa/Bj5I1oEtsDKHiulDZqnHzYPnj/AmptrML7ReFTyrmS0D/2HnFjWcfa+LXe3oEGbBnKHrOXW01v44eIPgsfaFr3N5L4BfmNLpVrJbwsl8LyUMoEUqAq0HkxXU7na2qF7hgq/ZBiRWUwVWDfe2WhUqClQF2Ds/rGcMs09l5ydzFsOMJqd80nnOfcdwL339Mf9QqlbXmFHTAb0hAMTbWqkZNdutq4Z1vdaj8b+jQXrSBVq8pX5nHtcTBgT+k6BpbiR9udfmI92we1wJlFYoLQE5jgBWJKiFRGumDP1+FSo1Cp8ePRDk9WdR+OPSq5b2rU05l+Yr91mX1Sx6bEmHd8Y7Hgt9kRMU2Ns4tf8NnzagAkHJuBQ3CGDyKxCvyffxJSQlcCxtZDzW4jFNVGpVRxNijH+vvs3ACa2jynwPZhG7huJLdFb8N5/70nqQ//8sH8bMe0JJ8WBAM9zn2PWmVm8ARrZBtR8x+LzxpEDX8BGlUrFaxBrTt4jMU8VIR6kP7BrBHO+307oHOgLvW8dfMuwrYRnaUZ+Bvrv4AaEZbfbeGcjdj/YbbQffdhRqcX4+NjH2s8JWQkGS/VSNBgx6TFotq4Zwv/RmWGYMo/oa8nW3lyLtw+9LUnQMEdTY2quK0tDQo2VMPXiEGrHp/kRM8I0lwJVAS4kX0CeMo/zgGyytol2ycUWRD6OlFSPPUZNzAntPp4QA3LhW6LKKcjB4stcj43ErET03NIT7Te2Nzi+JVS7kjQ1L49n7tsS38SoeXPVCF/GvpP+Q5ltrKoPO8aJlBAI3577Fv/c/QfD9ww32Hc//T5nW/+7uDjJVlJz2ysM2yvVSsHIxfqo1WpJ+eCkGLHrc/XJVaNCjWZM+2L2aYVfIYQyRStVSnx95muDcj7j1HIe5QzKeMdlRNMlFDE5NcfQgJz9W+y4vwPTTkwzqCOH+8911xTbZV7/WD239MTIvSM5+6UIJ69uf9WgTOwlkv28Zz+f9G1sNEhJVmnO8pGYRs2WkFBjJUy9OITadajYwWhbzo3DuqbkuINPPDQRO+/vxJLLSzB2/1h8cfILgwt04qGJKFAV4Paz25In6h33d6DR6kZIyErAlSdX8OnxTzlW+kJICe6k/wBnPwh+vfIrfrvyG2e/3Dfn/+L+wzuH3zEo1zceXHJ5iVYI43uI8Xk5yUGlVvHaYRx+eFh2dl21Wo2jj44aPJzZGJv4lSolhu8ZLhqFVew+0L+u2OfTmKHhnWd3sC92H/8x+a5JvWes/nfLU+aJHo/NpceXeHMsCdnU8JVNPT7VbG0RwAjRd55xY4YooDA6qTzMeAilSolPjn+Cr898LXgd/H71d3TYxP/sOfDwgFGBSEOIdwj/7yLh8SEmEGjspvgEeEu70r/272vaftkCDmBcM2HqWDSaWbVajTOJZxD5OBJv7HgDf1z7g+PlKMU+S4rnoBpqpOel42LyRdljtmeeOjbmva4QgpiiNixUFQouVUh5eMixOK/qUxWxGYbHOplwEicTTmrfRPfG7kW/mv0M6o3fPx6XUi5hctPJGNeI+4B/nP0Y0c+j0S64nfbinnGSWYfvuaWntt7T3Kf4o7tp7qz6NxD7JtJMpFn5WbwZvOUaDS67wm8Po6+9+f3q75jTXjiLsFjm5jxlHtyd3Y2Ohc9QePLRyZxtjZAnpu04HHdYmx5AyC7DmGB+NuksY+eRKvzQFnvY6gujbM8rY6y5uYa3vFBVaDDJ8x1L/6Gbr8yXdP4BJmM0H89ynwnGF9JHSCCT0pZNjy2GbvFCE4q3m7d2WXT+hfkcg9L/Hv2HYfWGGbTRRM7V4OakM4hlB8A0hkqt4hW8xeIYAcDO+zsNXh74tBd8AjjbM02DlECwxvjlyi8Gy8CJ2YmIfByJZgHNeNvwzQfpeelYdmUZ+lTvg4blG/K203zXU4mnOPFh9IPfWUqAU6vVGLxrMBKyEjC3w1z0qd7HaJuolCg0qdCEGw6Clp8cDynaAP06353/jjdabVRKlKQHyM2nNyWPr6xHWfEKrPue79ia2Bcrrhm6Oob/E46Jhybit6u/GexjI+aOKRc+dbuQp9C+GGkTiljfAP/SgP4bnFRab2ht1H5CSFMjhFjsnNOJ4nnWAOMPykmHdR5BQoKimKBiqjbzee5zQW+u2WdnY8juIQbl+vZT5nhDCWl1vjv/nSzvJzZCSTVNDZPPNw7965Wdh4xvyYIvhEOv6r0MyqSgVPMHJjT2gvHZyc8MyvjOv6uTq0EZO3yABinXnNh1/+D5A/x65VfefaP3jRZspwJz7/5w4QecTmDuvR8v/oj1t9ZrPcv40JwfY+lxpITbkLJ0rYZaK3wejBV+CWMzah8TVoVjU0OGwg6IhOe1/s3D92bx9MVTAxsRofaWhN231MSJSpWSY2/DpyVhw/dwEtNmcManl8fGWEh8Nhcfy8ufJfRmx/d2aGo8i0JVoYF3iz5qqCW5dP9+9XejdaRkpjY2oUqJAyImPEU+jsQf1/7QaielTuCLLi0S/I23RG8RbHcvTWeEbaoh7d4YYZug+Kx4kxM36gtd62+tR/d/upskJCsUCt5x6Gs4+F5I2OjbhZhDgaqAo6WVUl8oiCSfYH/g4QFJ/Uq5xsQEH3ZoCDmo1Wr8c/cfrL65Gm8dYoyh2d5ay68u522nEWqM2V9J0dLL1VBLFUw059QWaTikQEKNlTD35tGw+PJieDh7mDUWvrdaY8dm3wBiavnMAp02oOm6pmaH0JaaJZeN3AlKrjDImzdIxBZFnzxlnuASFhtjAotKrZL0YFKqxQO1nYg/ISsRXXxmPBqtbiSaZ0roTbKKj3AcqcNxh/HTpZ+0BpK3nnFTEuy8v9PABRjgN9qWQtSTKO1n/YeuVIFq6vGpgvuSs5N5BQEpWttlUdzrY975eUjKTuJNhGsMBQSEGr1rhy0c+HsaxlHiQ4pQHVwq2KCMz6AY4D/vCyMXounapoIednz3ANsLVAwpz1wpAr9cClQFSMjkLr+xx6LvcKBBqVJiX8w+3phWbOZfmG/05VPKd3/6QrdkrVAokFOQI+l5ueHWBk7yVHt635FQYyUkLT9JeJBm5mcKBnaSrL7nqWbJ6MCa/Di2jHjKuYEF0iQInR+pmprotGj8ef1P3Hh6w2CfHNfqlJwU3qi0+kg5f1LftvjcV5UqJW6k3uAsG4mhOX8RWyMAGLpJs9G8feqTmJUo+Tj6AsxnJz9Dv+38b8bmvgnqZ+Y2du6lxCsR4l7aPaM5pDRB6ZKzkzl2LCkv5AtwTgon3nteTCDhW77hQ0oCVT6tgtCEy3feNZG2hTDrOSPhkSkkYJjD/tj9ojGbhChUFWoDnxrjbCJ/mgk5LL+m0xjdfHoTrTa0wswzMzH/wnzsvL9TUNCce34u5ze25/ITGQpbCUsmtBRy6xSThtn7MgsycevpLdT0q4nFlxejVVArSceVyqnEU6hXrp5F+5SD/vJTTmEOLqdcRmVv+QkT2ejHvdA/pqUxdj2o1WrJcYI0Nk8app+YLjoh7Y3ZC08XT06ZJYTUb899a1b7nEJDO6MTCSdwIuGEyX2ysydrMDbBjN0/FlteFV7aEkMT/fbkkJPwdfflraMRON45/I5BsEi5qNQqXg2XpTI+p+WmoYxHGcH9cqJBm3KNmRMrS/Oyma/MR3ZBtvZ71C1bV5vWYe3NtZjaQlgjZwrebt4Gy+1Snv9yPORmnpmJLpW7GK8oEY1tzdborbLbkveTAyLFDsWcoFwA/wNfiEG7BmFqi6lYdWMVVt1YJes4xh74YqHvJx+ZjB87/Si4/0XhC4PJVC4KhYLjtj7rzCwAwPth7wu2MTfK8rpb64yqhOVi7DyrIG35CTB88IsJNKP2jjIQggDbZpf+9PinkuK2WAK+xImayVXI7fVu2l2Oat4UkrOTRYWatTfXmizQlHYtrT1/QvFYRPMIyXiWJGUniQo1cjDlGWiOsK25pntv643k7GQceuMQAkoFCMbRqeBZQZa2TChqb1mPsgapbvSXW/nQ93ISQ443mrUh7ycHZMTeEUbr/BL1C7ZFb8P6W+t53VA1mKIV4JuQxIQPuX2xETPOPBx3WPStuuX6ltrYLsceHTNpfELnZ9s94VD4W+6a9tatYf2t9Wa158NYEkc5mho5yyV8Ag1g2zwue2L2mN2HMe+xZVHLcC/tHm/UZs1EKbZcZmpkZg1inmtPXjyRbBfCh5w4O3zICQdhyWVmUwRnczRO+ukawv8JZ547Ao9YufeAkLv//AvzTfqucl++xDwebQlpahyAG6mGdhdS+PL0l0brmHKB8L1dmHqhXU81nvVZ7EFn7IE5et9oXBl5Rauml4uQUCM2Jjmu0bbC2KS2+sZqnEo8JakvSwgk9swKbQpTjokbmae8SMHrO14XrSMWwfer01+ZNC4N1rzmbHk9W2oZCzDtGuNr4+niKdlLU593/3tXMLhpvsryRsPWZMKBCfYeAgAyFHYI+GJjWAJLLgGcSpA2Ieqz+e5mo3X+ufuP4L4vTn1htL05D0ohYU3s3BW3CRuAZIEGsMx1U9zOkanXN6D7rmLCoJwlGj6kxAkpDljyujDlvtdvczftrmSBRq6wbw1PKADYfMf4M9UU+JwabI2UiNbWhISaIo4llwD4Ighbim/OfmNWe2tMoGLnTo73UnEku9A8myE2bYPbWqwva2GOdxLAEmqsaEdU1N/6NWlLHmU+Es2TZMl7NS4jznglPQ7HHeZsD9gxQHJbobEL2faYu6zHZt2tddrP5j4vizK2XLbmg4SaIo7cnD7FFXPesoVUnWJvgUUl27i1MNf+A9BNAFKiD9sbcwV2jTBjzQeylBgv9mTwrsEAgF5be4lmtN4Xsw/zzs8TTJwohw23N8hu89OlnzjpGuQgpHkREmatqa1sF9zOan2XZMimpojz9MVTu65P2oLmAc3xwZEPTG4vNBGJBWmTktytpPMw8yEarW5k72HYBBVUOBx3GP/F8UfvtgRawcmGXmVyeJr71KixOsAf+dzWmKr16ru9L06/aSik6/8mydnJstLOmIJQ/DHCPEioKQbYc33SFpgbqMkUbZa+CpswxBztma0x115FqVJi8pHJlhmMABrhuyjbKpkarbm4IOQdpL/81O2fbrYYDmEFaPmpiOPoWhrAurY+QhTliYWQj9zYS/roZzq3JueTz9vsWHKxZyRYe2IP7dmRR0dsfsySQMm8gosZji7Y2OPtsCi6dBOmcyzetBhHGmLSYyw0EmE0mhqhWCZFgZIq1NBLjuMg+wo+fvw4+vbti+DgYCgUCmzfvp2zX61WY+bMmQgODoanpyc6d+6MGze4bmZ5eXl47733UL58eZQqVQqvvvoq4uONB4b75ZdfUK1aNXh4eKBZs2Y4ccL0UOnFBUvGhCB0mBrTgiBMpThMnC4Kx7dI4NPKpOeb5z1HFB1kCzXZ2dkIDQ3F0qVLeffPnz8fCxYswNKlS3HhwgUEBgaiW7duyMzUrWVOnjwZ27Ztw8aNG3Hy5ElkZWWhT58+UCqFJ/BNmzZh8uTJmDFjBi5fvowOHTogIiICcXHyXQKLE+U9yzu8TQ1BlASKwwvKzWfWNY4VonOlzjY7Fp9jQXRatM2OT1gX2UJNREQEZs+ejf79DZP9qdVqLFq0CDNmzED//v3RsGFDrF69Gjk5OdiwgXHdS09Px4oVK/Djjz8iPDwcYWFhWLduHa5du4ZDhw4JHnfBggUYN24cxo8fj3r16mHRokUICQnBsmXL5H4Fi2PNN7DsgmyLuE4SBGFfTInJYmu+PvO1XY5ry2Wv4iBcEqZj0SspJiYGycnJ6N69u7bM3d0dnTp1wunTjBtdZGQkCgoKOHWCg4PRsGFDbR198vPzERkZyWkDAN27dxdsAzDLXBkZGZw/a2DN+BPxWablayIIomix5PISew+hyGJLo1lb2E8R9sOiQk1yMpMkLCAggFMeEBCg3ZecnAw3NzeUKVNGsI4+qampUCqVov3yMXfuXPj6+mr/QkJCZH8nKfxw8Qer9EsQhGNhLOlmScWWUWjlRCAmih9W0fnp24Co1WqjdiFS6sjtd/r06UhPT9f+PXpkPLCUKZib8bm4UKl0JXsPgSCKNZ8e/9TeQyAIh8aiQk1gYCAAGGhPUlJStFqWwMBA5OfnIy0tTbCOPuXLl4ezs7Nov3y4u7vDx8eH82cNSsoabb1y9ew9BIIo1hyNP2rvIRRJulWhYHeEZbCoUFOtWjUEBgbi4MGD2rL8/HwcO3YMbdsySfGaNWsGV1dXTp2kpCRcv35dW0cfNzc3NGvWjNMGAA4ePCjYxpaUFKHGWeFs7yFYFXqwWo6yHmXtPQSiGOHq5GrvIRAOgmyhJisrC1FRUYiKigLAGAdHRUUhLi4OCoUCkydPxpw5c7Bt2zZcv34do0ePhpeXF4YOHQoA8PX1xbhx4/DRRx/h8OHDuHz5MoYPH45GjRohPDxce5yuXbty3ManTJmCP/74AytXrsStW7fw4YcfIi4uDm+//baZp8B6TGg0wd5DsCguTo4dw+KdJu/YewgOQ1EJ4mZLV2HCdBz92ULYDtlX0sWLF9GlSxft9pQpUwAAo0aNwqpVqzB16lS8ePECkyZNQlpaGlq1aoUDBw7A29tb22bhwoVwcXHBoEGD8OLFC3Tt2hWrVq2Cs7NOE3D//n2kpuqSDg4ePBhPnz7F119/jaSkJDRs2BB79uxBlSpVTPritiDUP5Sz3Ta4LS6nXC62gd+KykRlLRw9crMtKSrn0s/Dz+S2g+sMxqY7myw3GEKQE/GOH0iVsA2yZ6nOnTtDrVYb/K1atQoAY8w7c+ZMJCUlITc3F8eOHUPDhg05fXh4eGDJkiV4+vQpcnJysHPnTgPPpNjYWMycOZNTNmnSJMTGxiIvLw+RkZHo2LGj3OHblOp+1dGvRj/tdlCpIJx+83Sxjdrp8EINBTl0OMwRrj5v/TmWhcuLgxXgJWzjV1JpXL6x0TppeWlG6xRFPm1hf8PvCp4V7D2EIoVjz1J2JsQ7BP5e/tptlVoFFycXsyfPxv6N4eZk+7T1ReXt21o4+vezJbZ00bUGGvuq9hXby2rXOaQz6pSpY40hFVuevHhi7yFYhS9af4EhdYcYrfd2qHVNJCyxdNeneh8LjKRoQEKNlWEbEVsi8vAPnX7A2oi1aOTfyOy+5CJVU7MmYo3Fjz2zzUyL96mPse+387WdVh9DSaB5QHObHcsSwtXmPpstMJKSyd7+e+Hn7mfvYViFQXUGSRIoWgW2suo4elTtYVK7Cl6MhueL1l9gboe5lhySXSGhxsq0CGih/ax5wHo4e5jVp5PCyS7J8TLypUVkDqsQZvFjd6ncBRt7b7R4v2yMaWqq+la16vE1BJYKtMlxrEnqi1TO9metPtN+bhbQzNbDkQ37WpATysDLxavYa6ksSSXvSojLNJ4ewpG1pM0DrSvER1SLMKndrtd34e++f2Ng7YGS6n/R+guTjmNrSKixMmz1tUYQCa0QKlTdKJ4ungDs40Z+8OFB45WshAIKNCjfAIu6LLKa3YIlbWpeCXkFS14xLSx+Rp510nnYkzfrvqn9XKAqMKkPU8+nEGMajhHcJ6a1q+ZbTXBfeJVwEmr0yC7INlqnUXnLaJ4b+zdGRFXTJnlLUr9cfQCMCYIlWdljpUGZqbaOni6eqFu2ruTnXmnX0lqtkOb7FUVIqLEy7AtGI9SYY1zWLrgd88HCz83gUsGW7VAEHzf5QRA1N27Xyl2tdkNZUqj56ZWf0DKwpUltcwpzsLDzQouMoygarpoi1IR4h5j0Nq9WC98oU5pNEdwndqyG5RoK7qtT1nL2NEXJoYAtlFqDIXWHYEX3FUbrGbsG+lTvg17Ve1lqWJJY+spSg7JpLafhi9ZfYHXP1RY9linPTkuhhhqz283GD51+wPLuy+02DmOQUGNDNEKNqcsYA2oNgLMT4/ZuqbfBVT1XoWVgS/zc9Wcs6rLIIn2KsbHPRvSqJu+h06d6H/i6+1rk+FqhkAcpk+arNV6VfCyNVk2DnDg44VXCjVeSwK7Xd1mkHzF+C/8N9crWE9VgsClQyhdqOlXqJLuNBpMCu4lcCuU8y2Fdr3Wo4mMYTsKSASrLe5U3q/13Hb6z0EiA98Pel92mum913nK+wIx9qvdBy6CWRp8NHSp1EN1vj2WsTiGG16aXixcG1RnEcRQxl25VuqG6H/ecVixd0WL9syntWpq33MPFAz2q9rCrcGUMEmpsiLl2MGw1o9gbqByaBTTDih4rULNMTXSt3FVWW1Me4A3KNZCtEZnZdiZn21oPLilq3NplahuUTW0xFXPazzEw2NP/nvawlfFw8cC+AfvQJqiN1Y7RNKApNvfdLHnNvbG/cRdffU4nnpZ03dT0q4mfu/6s3fb38kdNv5qyj3cm8YzgPpVahVD/UHzY7ENtWb2y9dAqsJVFhRpzvajaVRQW4G3Btn7bDMo+bfEpr22c5rfdE7NHtE9j935RCTthjSXI+R3nGwjoK3ustEooiv61+hutM7XFVO1ntjF4OY9yFh+PHIrGFVBCMPdCv5Z6zWJ9mQJ7YqzsXRmLX1lsk+OaMlEIxW4w97zxPTTLeJRB3xp94e7sLtrWXlFTK5auKGo/Ygpft/1a+1nzUG0R2IK3rkZjtef1PZjXYR56V+8t+3gaDaUxXJ1c0bFSR/zU5SdEVIvA+EbjUbdsXYN6xjz0nuc9F9ynfTlhXUob+2zE8u7LoVAoLPbC4eXqZVZ7qddb66DWZh1HCL57ZXj94QgqHWSV4wHM718cDNFNQfN7ftTsI21ZcGnLmw3ULVtX0jOX/bLA1qR+1eYri49JDiTUWIHpLadjfa/1BuVsTc2WV+Vn9r797Lb2sz2EGvZDcmbbmSarIOVqWvRvMClvJoXqQoOyAK8AUW2ZlLc8vrE7SbyN7BkK3tJ5u8p4lNF+NvZ7alxHQ3xC0Lt6b5PepqXal2jO8SuVX8H8jvNRyrUUPm7xMafO952+t7iHnpPCyfJvzKxbfGjdodrPUpfTvFy8JE3wr1R+xWgdOd+tcfnGGFLHePwWKcg1tO1RtQe83byNV7QwO17bYbG+Xq/5Oq6NuiaYP612Wa622Bqaa757VH/OYW+z61f2qWzx8ciBhBor8Hqt13lV7OwfXmi9WSqWehuUdUy9i9pUg125D39TJgs+gWt+x/nix5HwcMgpzDEo0/yufL8J26vDHgETNVh6wmVfy3znTUoUWbnHkyJE831P/XYqlXnLwMYSn1oq3AL7fmNrqra+ulVSe4VCgVU9V+GP7n8I1pnbYS4G1xls+iB5WN97PWa0niFa5/War0vqS/+e+qDpB7z1mvg3wduhb5ut3TKVar7V8GOnH7Xb5jyfNdfPoTcO8e5vE9QG74e9r/UGNEWo8XY1FPw+avYRXJ1c8UXrL3iFGjGng2YBzfBB0w8wvN5w1PCrIXs8loSEGiugf0F80vwTBJUK4nhbGHtz7l6lu1XGJgcvF+4Dol5ZbrwON2fTJun0vHSTxySE/nLTuEbjDOo0DWgq+gCQMq5t0YZ2AprfWwXDyYz9kDU1PlHHSvzpQEyxTbEUnPPI+qixAxlRfwR/XRNxcXJBqH8oRtYfKW67I2EuORx32KyxNA1oKrpfrvHm6TdPG95begKwtTyh+lTvI1lzZiljfYDR9H7e6nOj9diCXeeQzqhVphZvvbW91lotIa3U5Tm2EbM5mnRNW1dnfo2cQqHAhMYT0DmkMwD5S1Db+23H4UGG98DohqNxfth5NPZvzLkmJoVOwpRmUwyDZrK+Yt8afTG+0Xh82tL+aSNIqLEC+g+JkQ1G4sAbB1DJu5K2TOzNeWzDsSjtxm99rsFawffYHh2b+mzCpCaTtNv9avbjayKbXQ/M88jhmyTX917PsfMo5VqKv63Iea/obXwyyizINCjTCjU8vwnbzsZUIVBIKHj64qnZfZiKkKZmTcQabOyz0eQop2IoFAp80uITDKoziNfzSCr6gQEtzVdtvkLPqj0l11dAgQ29NxiUsydGuZo2IUHYVNyc3LCv/z6j9YxpsTQ4KZx4DecH1R4k2EYoEvXGPtYLytm0QlPeaLtNKxgKtuz7wJz7jS8Gmdj1bkw7pa+prOFXw8AzU4Nm+Zb90t2xUkeMaTjG4BoUWn6yN0VnJA6EVBsLIZwVznaLsLmq5ypMaTYFu1/fjaq+VTExdCJODjmJfQP2wd9T555oiluuFAK8AnjjPhgjsFQgXq+lU2mbov6Vcs75li60NzvPIT9urrPncHFywb/9/sXfff+WPTY+ErISJNe19PIT266Dfd68XL20Hm4a99z/Nf6f2cfT/222vboNJwabltlZ84ZrLkKpSgJKBeD7Tt9L7kehUBjYW6mg4lzD7EmGbc8EGGpQAWBxF50RP3vyMTVkv6uzq+iLVk2/mtjcZzO+6yjdjZxv8uaL3ryq5yqMazhOGyvn5JCTnP2VvU2z4WAnGxZidcRqlPc0dK2/lHLJoIx9j5mjqWG/HGnCJJiTm8mUtCRFSUiRS/EdeRHG3AtiZP2RRuvoG8OtjViLiaETzTouAJT3LI8xDcdwjL183X1RsXRFzk2rMf7UD5jVKsi8PCeHBh7ijfvARn+Cfi/sPYM6Qg8VMcFFbOIP9WeiQPMZIGsmHD5NTKXSOu2ck8IJ1f2q83rjyEUoJswrIfxGn5YWktnnSui8ze0wF3v770XfGn1l9T24zmCEVQjjeHnoT/quzq7w8/AzaCv0u7OvkZZBuqCI5gQ5rOBVAQffOIjTb57m3S/VWJb3t1Eb2tSs67UOK3usNFgG4jMGZtvgsIUjhUJhtTw/9crV4zViZi9FstGMW0wLoVar0SygGSY3m6y9v9gxVH7o9IPJhsH5ynzR/WMbjpXVn7kvsxrYv9fKHivxQ6cfeJfTjTGi/giUci3FCT0gFSlzmL1tZ4QgocYCHB7IXZ80963Yz8PPaJwA9lILADSp0ERSRuH2FdtrXW/bBreVNS72w1dz0bMnCACo5sNMtubaBMl5oLA1AZrIyELuxWJzu9DE/0nzT/BTl58AAEqV4dul5lxMbDIRlb0rY1rLabzHkyrsSonurG9X8mePP7Gxz0ZB12chIWh2u9kGZVV9qho9vhRVu5PCibPkKpXh9YZjTcQalPXUeX+Ya8/BvkbYcTQ0wrkxhIziA0sFCk6q5rgW6wtnCigQ6h9qcF37e/obfVZ4uOhsuV6r+Rq6VemGMu5lLBqZW2w5XEgj4uvui7NDz2J7v+3aMin2SOz7qIYv/8QqtPzMZm/sXoOyxV0W4/KIy9jcZ7OgUbI+SqUSubm5yM3LRZBbEILcgqDKVzFlL/805VL+SitKa9uVVpRGp8BOUOYrOf3p/+n30atSL7zf6H0ceu0QAt0DteUV3SuK9qP584CHto2ygP/Yfs5+2NRjE3b32S2pT2N/SqVlUv8UnTjcxZgKXhWwqc8mDN41WPBNWS7GjEBDfELQvUp3HHh4QKtFkIIaasxpPwfbordhYB1picw0cCZlAeFAU+ebdt+gbXBbzDwz06BOY//GuPrkquixpE42+ux6fRdyCnMEJ0FRTQ3Pvrpl62JkA53mjE9lrvnOFUtXxO7+u7n7BN4bmlZoyqvCBpiQ8fr9dw7pjKOPjmrL9L+fJmme0LJbGY8ymN1uNq48uYIH6Q8Q+TgSgKFmbUCtAWhfsT0+PCr97c5cIf77Tt+jcfnG6LGFscPRTOjs7yL3rZmPxV0WI6sgi2PLkZWfJamtScuZEs+L5ppa2Hmh9rwHeAWgV7Ve2nxrKTkpgu1rlamFE4NPYH/sfsw+NxtN/Jtw9jfxb4IhdYbAxckFvar1gkKhwOFBhy3q5i92X9UpWwdrI9by2tDoCx+tg1rj0xaf4rsLzDIWn+aNfV6FYt5s7L0RW6O3IqsgC3/flbbcW8O3BrpU7gJAWhJTtVqN5ORkPH/+XLv9aU3GUFb1VIWY5zHauppyKXi4eCAmJsZ4RRb6/VfwqsDpY36j+UjPS4efu5+kvuuq62r7VD9Vc74LGwUUyEY2YiBvvEL4+fkhMDDQrGcKCTUWon65+rgy8opF1yKdFc6iiStntZ2FdhXbGRWkGpRrgBtPbwAAbj+9jcBSgZjYRP5SlZQlDI22yMvVC92qduMVakq5GH+LkjqJsON3AMyyhK+ziW/1el9vWL1heKvxW5wyXqFGROEpdHNWLF1RUKjRnOdPmn+CNTfX4JPmn6CSdyVkFWSh7V9tOXX0EX1jrtkP/Wr2w9j9OgFBf2Kb2XYmziadFezDGvSo0gMvCl9otzUTWaFKt9RnidgjmgmLTVpemtn9CiH1WaBZVmHHZBnbcCxHC7oleotBZG1Ad678PPwwsM5A1C1X1yDqtUKhMHCx1l8mYj8j9JnXYZ7R72BsEmpSoYnRPjT9DK8/XCvUCHFk0BEUKAsENTJVfatiSvMpmH9BPIwDmz96CLu+86ERaCpUqAAvL8ZYV/mceT6EeIdwNGSFaYbL1mwqla6E+Kx4AIygJ9ejid2/t5s3ArwCzBIM0nLTtAb1lUpXgqcrv2GxpVCr1cjJyUFKCiO8BwWZHqCRhBoLYqpA82qNV7HjvvzgTaXdSksKZ702Yi2armOs9Z/mSveY0UfKTcL2IBKa7KX0I2Zox57Q5boQytHU9KvRz8Aokw8nJxGhxgSPCM11NLLBSIyoP0J7vtgTu9C4+NzK9WELjHzXrJRxmhv8cXCdwdh0ZxNql6kNhULBawMSXiVcKxRLvbfkalSkaitM8TY0ZmPxTbtvUKl0Ja2HHPu+8HT1lPY76P2WcrS2bIQC+s3vOB8R1XRZrztW6ojj8cdNOoYpJGUn8ZbzGe/yYSxQ4diGY3Et9Rp+6/abrBxhHk4eWoGmXDlmOVOtVsPJlfnN3T3cOUKNppyPBuUbAAAy1ZnIzM9EgG8APFzlhX9g91+1XFVZbflwVbvCqZDp06+0n1VSMejj6ckITikpKahQoQKcnU3TJJJQUwToULEDr1BjKeNOoXgHcpE7HvZE1C64nXa569OWn6Lfdsu4h8sVJMVuTv3vx1e3VplaiE6L5pSJTYxCxxMqr+pTFX1q9BGs92ePP5FTmCP4UJfrlcY3DmO/889dfzYpnxKbT1t+ij7V+2gf6HzeVGx3eKlGmHIfvpKFJROEuOaBzeGscEZgqUBEVIvAH9e4moBWga0El08alm9oUw9IofOmL8wNqzeMV6gxxRjVFoxpMAZRKVGCWlFTx+3rwmiDNRoawLRlWPaLSoh3CArVhaYlYLUw7DhJthBoNGjOZ0FBgclCjVUMhTMzMzF58mRUqVIFnp6eaNu2LS5cuKDd//jxY4wePRrBwcHw8vJCz549ER0dLdIjsGrVKigUCoO/3Nxca3wF22If723ZsCeAQC/d+rhQpF722/dnrT7TJswUiqbMNvpjJ0izJGIeYvo3L9+kMqHRBMN2FvwBd7y2QzA8OsBMlGIxSApUOqFGKNgfW2VvzKByUedF2s87X9uJqBFR6FipI8p4lMHBNw7i1JunRNsL4erkiiYVmmgf4E4KJ4xpOAZ9qvfhNWq21oNVak4pUzQ1vu6+ODLoCP597V9eo1MxgaqqT1WzNZqWQN+2jc+5oJxHOYvHxbEUfh5+WB2x2uL9aoRsc69Ldxed4K5QKIqEQAMwwlZgqUBBBwNrYYn73CpCzfjx43Hw4EGsXbsW165dQ/fu3REeHo6EhASo1Wq89tprePDgAf79919cvnwZVapUQXh4OLKzs0X79fHxQVJSEufPw8O0KK1FCUslX7T2m51CocCxwcdweOBhTsAnoeze7LdrY99ldIPRWBa+TLsdUS0C/Wv1x7ftv+Udh6k0LN+QEyBME0dmRqsZks6fsaSVYrCNe4UmNHNvarYdisZjSx92yAAnhRPW91oPVydXfNL8E4MxsJcTQ7xDOEJAYKlAk/N/8TGl2RTM7TBXe3y2G7e1hFz2JGKNgIFlPMoIXjNiQo3i5T8N7HhHgC5dwKy2sywwSmGkxDixRq6lgbUZra5UDyRrY06wR0AnHOovG1vKDdzSKBQKlPMsZ7e0E+Zg8eWnFy9eYMuWLfj333/RsSMjvc+cORPbt2/HsmXLMHLkSJw9exbXr19HgwaM6vmXX35BhQoV8Ndff2H8+PGCfSsUCgQGGlrQF1cWdV6EmIwYwbDrId4hiM2IldyfLZJc8mkRxFx6NehHsKxXth5uPbul3f6o+Uec/S5OLlZ7YM9sO1Ob7O/1Wq9jaL2hVn1DWth5IdLz0jnuzdY6XqeQTlpVu5Bbeyk3nXZGAQUa+zfGxeEXeSdZdqoMWwfkcnFywcbeG1GgKhAM/PZ+2PtYfFkXaE6uYN+wfEPt54G1B2J/7H7eetaK4M2GHdNIo4nWoP/GPL7ReAyrN0wwMqxc9M/byh4rhcMisHB3dsc37b6xyBjYfN76c4yoP0JSeAEprIlYg8jHkfjpEr+gb4yVPVZi853N+O3qbwB0Qpc+rs6uKFQVGgiy/l7+KONRBi5OLkjL1Rmni2ll5VDJuxLiM+PN6iM2NhbVqlXD5cuX0aRJE4uMyx5Y/ClVWFgIpVJpoEHx9PTEyZMnkZeXBwCc/c7OznBzc8PJk9xIkfpkZWWhSpUqqFSpEvr06YPLly+L1s/Ly0NGRgbnryjRtUpXjG/EFeLYb6eahGXmwM43ZTUE5hFnJ2d80foLfNz8YwM1NnuZhK2hkXY4aRPXrtf50zH4uPlgdvvZ2gjEbAFDo60ApGWqFSpjE14lHANqD+CUTQydKDtHkBTaBLXRfpai9dGcS6G0B5V9KmNcw3H4qNlHNl1b19CgfANRz5lh9YZJTo7IR1mPstjbfy+ODT4mWs8WQo2Xqxf+G/ifNlIy+3fgm/wsJdAApr8QnRt6TrJnkxycFE6o5lvNYtdcWIUwg2etHCp4VcC7Ye9qt8t78du01fKrhbpl6/I+O/QDSFbxqSJ5+dMYpV1Lw0nhZNFrorhicaHG29sbbdq0wTfffIPExEQolUqsW7cO586dQ1JSEurWrYsqVapg+vTpSEtLQ35+PubNm4fk5GQkJfFbugNA3bp1sWrVKuzYsQN//fUXPDw80K5dO1FbnLlz58LX11f7FxIiL429PWDHmKjqWxV7+u+R3JZvsh9eb7glhmUyg+oMwqgGowzK2cskUoIGmkIVnyo4MugIhtYdii2vbpHUZmg9nYu4NTUT/l7+2Ddgn9HJVC7s8ypkwMzxyOIzFNYrm9xsMkY3HG2ZAVoYL1cvfN3ua+MVRajkXQllPcraRNNpDH8vf22kZPbvJ5Yh2Z5YalK2FezlUnbaF0uhUCiMPjeq+1ZHcOlgo/n95ODs5Iw6ZevY3AamKGKVp/batWuhVqtRsWJFuLu7Y/HixRg6dCicnZ3h6uqKLVu24O7duyhbtiy8vLxw9OhRREREiFo7t27dGsOHD0doaCg6dOiAzZs3o3bt2liyRFibMX36dKSnp2v/Hj16ZI2va1H0JyL9bL1isG+mTX024Z++/1jM88nSfNKC0YiMaThGdls5Swzl/9/evQdFWbZ/AP/uwu6yrLAjJ2GBJTRfUUEkFFT4ZYiDB7TXrDRF0TezsUIgZhSNEstIqzHNPFRM47yNpo2pjfk6KpqiDgTJoVA8jgoKi5Ihi6kgcP3+IJ5YF5FlT7Bcnxln5Hludu/nWp7da++j3A3Lw5frrd3Rmce2RHeLi4MLwjzDnlywHY/uog7o9tkb0lLzpGO9QUfTwc2R8LQdJNoeiZ0Ey8OWY1nYMrg7mv4DuK1H12OyVW2X9n+05cQQxrTcySXyTi0VYSixSNype765uRkff/wxnn76achkMqjVamRk6I9dBIDs7GyEhYVBJpPBy8sLy5YtQ2PjP1+cfvjhBwQFBUEul8PV1VVvbOzWrVsxePBgODg4ICAgAJs3bzb+Qp/ALFO6BwwYgOzsbPz111/QarXw8vLCzJkz4e/fkkWGhoaiuLgYtbW1aGhogLu7O8LDwzFiROc33hKLxRg5cmSHLTUymQwyWdcHdlqS2kmN8rpyTPTX3d3XkDfTttNDH7cEurED3p6ksx+Iz/o8i9xZuV36ttLe/kumYie2wyT/SdDWa9vvz2/n5SjTlnVq/MHjjFGNQX5VvsFNx4uCF+Gzgs90jvk4+WDN/61BX1nn3jSt0aVkVkZcjqVbajozyLpty6E5RXhHWOR5rG3lmJXCchLGJDXtdkMT6SwiaSlye7lB9/Hy5cuRmZmJdevWITIyEhqNBufPn9crV1FRgcmTJ2P+/Pn49ttvcf78eSxcuBAODg5YuXIlNBoNZs2ahU8++QQvvPAC6urqcPLkSeHLQWZmJtLT07Fx40aEhISgqKgICxcuhEKhwLx5+q33pmLWdWoUCgUUCgVqampw6NAhfPKJ7tRfpbJlNsilS5dw+vRprFrV+QFnRITi4mIEBbW/U25Ps2PKDpy/fV5Y7r5VfVN9px/DxcEF2ydv73CGzlifjjeLtKSuNr/OHzofWWVZnVp4sCseN0X9cdrbJdkQ8UPj4SZ3M7jF5nGvc2z/2A5/70mJZ3vL2fcKHeQ0XdkmgXU/bWeadiWp6efYDzfv3URov1A8uKW7nMj9xvsI/864DX27Im92XqdnKdXV1eHzzz/Hxo0bhcRiwIABiIyMxLVr13TKbt68Gb6+vti4cSNEIhECAgJQWVmJ1NRUrFixAhqNBo2NjZg+fTr8/Fq+LLf9PF61ahXWrl2L6dNb3qf9/f1RWlqKr776quclNYcOHQIRYdCgQbh8+TKWLFmCQYMG4T//aelq2LVrF9zd3aFWq1FSUoKkpCRMmzYNMTH/bIIYHx8Pb29vrF7dsqPs+++/j1GjRmHgwIHQarXYsGEDiouLsWnTJnNcgsU5S531NodsPd6qM9Nan7RnlDm+jdqL7NFH0gd3H941y+DXRw1zH4bsmdlmm+bboXbyAV9n48ZqScQS/Ptp0yxGaAqqPipsit5k0una5vavvv/CxZqLmNJ/ypMLP0ZH90Z3GG9jbjun7MQr+zu3s3hP1bZLebx6vMG//7/p/4O2XgsnsZPJ9juypHPnzqG+vh7R0e0vw/Fo2dGjR+u0AkVERODu3bu4ceMGgoODER0djaCgIEyYMAExMTF46aWX0LdvX1RXV+P69etYsGABFi78Z22vxsZGoTHDXMyS1NTW1mL58uW4ceMGXFxc8OKLLyIjIwMSScv4Do1Gg5SUFNy8eRNeXl6Ij4/He+/p7jpcXl6us/z8nTt38Prrr6OqqgpKpRIhISE4ceIEwsK6Nh6hp2jb9/rNhG+MfjxzjJdoXb+miZqEfWzMzVRTIQ3W5rMt66Us1DfVW+3Dv6tdR53ZV6a7Lqb2OP+d+F+c+/OcSXfGbstd7o7rdd1/TJ4xhroOFf7fdh8qW9L2nnm0q78zZHYyuDu6t7voq9xejrzZeUbVrysM6bZu3YqgM4hI7z2mtcVSJBLBzs4OWVlZyMnJweHDh/HFF18gLS0NeXl5wsrAmZmZCA/Xbb3q6krBnWWWpGbGjBmYMWPGY88nJiYiMTGxw8c4fvy4zs/r1q3DunXrTFG9Hidvdh5uP7htkjcaU+7M25alkhlrazvg2NrdNF3t9lLKlNj/wn6jFhLsbvpI+xg1rulRrg6uOj9nRGbgg9wPuu0sMFP58d8/4n7jfav/bZtL25YaUy98JxKJuv1idQMHDoRcLsfRo0c7XBMOAIYMGYLdu3frJDc5OTlwcnKCt3dLi7xIJEJERAQiIiKwYsUK+Pn5Ye/evUhJSYG3tzeuXLmCuLg4s19XW7z3Uw/gKHE02c0S7NG1De9YC19nX+yI3WGW2QuGGu4xHBuiNkDtrDb4d809YLwnajtu5uuYr3XO+Tj56B0zhrF7Z5lL29lBtshe9M9HXtsVvnsLBwcHpKamYunSpZBKpYiIiEB1dTXOnj2r1yX15ptvYv369Vi8eDESEhJw4cIFpKenIyUlBWKxGHl5eTh69ChiYmLg4eGBvLw8VFdXY/Dgli9bK1euRGJiIpydnTFp0iTU19fj9OnTqKmpQUqK+dZP46Sml9j/wn6c+/McxvmOs3ZVery2q9BaW5Q6ytpVsBltu586uwRAV5mrxZR1zE5shy3jt+BO/R2zT5Pvrt577z3Y29tjxYoVqKyshJeXFxYtWqRXztvbGwcOHMCSJUsQHBwMFxcXLFiwAO+++y6Alm2LTpw4gfXr10Or1cLPzw9r167FpEktu7q/9tprcHR0xKeffoqlS5dCoVAgKCgIycnJZr0+Tmp6CT9nP/52zlg3YektJ9g/zLXYZ08hFouRlpaGtLQ0vXOPzvIbO3Ys8vPz232cwYMH4+DBgx0+1+zZszF7tmWWJWjFdxZjjMGy07Y5qWHMPPjOYowxWHbgd0/ufvo86nM4S52xOdr8q8MyZijufmKMMQCDXAZhVcQqeCm8nlzYSD25pWacehyifKNsbzVqZhM4qWGMsb9Ne3qaRZ6nJyc1gA1ur8FsRs++sxhjjDHG/sZJDWOMMdYFzc1d362b6TNFPLn7iTHGLODN4W9ic3HL4FruvunZpFIpxGIxKisr4e7uDqlUyq+pEYgIDQ0NqK6uhlgshlTa9RXqOalhjDELeCP4DSGpYT2bWCyGv78/NBoNKisrrV0dm+Ho6Ai1Wq2z76OhOKlhjDELM8fGssyypFIp1Go1Ghsb0dTUZO3q9Hh2dnawt7c3usWLkxrGGLMQP2c/lGnLMPEpw3eIZt2PSCSCRCKBRCKxdlXY3zipYYwxC9k+eTtKb5ci3Cvc2lVhzCZxUsMYYxailCkxWjXa2tVgzGbxlG7GGGOM2YRe1VLTumGdVqu1ck0YY4wx1lmtn9tP2ni2VyU1dXV1AABfX18r14Qxxhhjhqqrq4NSqXzseRE9Ke2xIc3NzaisrISTk5NJF0rSarXw9fXF9evX4ezsbLLHZbo4zpbDsbYMjrNlcJwtw5xxJiLU1dVBpVJ1uI5Nr2qpEYvF8PHxMdvjOzs78w1jARxny+FYWwbH2TI4zpZhrjh31ELTigcKM8YYY8wmcFLDGGOMMZvASY0JyGQypKenQyaTWbsqNo3jbDkca8vgOFsGx9kyukOce9VAYcYYY4zZLm6pYYwxxphN4KSGMcYYYzaBkxrGGGOM2QROahhjjDFmEzipMYHNmzfD398fDg4OCA0NxcmTJ61dpW5r9erVGDlyJJycnODh4YFp06bhwoULOmWICCtXroRKpYJcLsdzzz2Hs2fP6pSpr6/H4sWL4ebmBoVCgeeffx43btzQKVNTU4O5c+dCqVRCqVRi7ty5uHPnjrkvsdtZvXo1RCIRkpOThWMcY9OpqKjAnDlz4OrqCkdHRwwfPhwFBQXCeY618RobG/Huu+/C398fcrkc/fv3xwcffIDm5mahDMfZcCdOnMDUqVOhUqkgEonw448/6py3ZEzLy8sxdepUKBQKuLm5ITExEQ0NDYZfFDGj7Ny5kyQSCWVmZlJpaSklJSWRQqGgsrIya1etW5owYQJt3bqVzpw5Q8XFxRQbG0tqtZru3r0rlFmzZg05OTnR7t27qaSkhGbOnEleXl6k1WqFMosWLSJvb2/KysqiwsJCioqKouDgYGpsbBTKTJw4kQIDAyknJ4dycnIoMDCQpkyZYtHrtbb8/Hx66qmnaNiwYZSUlCQc5xibxp9//kl+fn40f/58ysvLo6tXr9KRI0fo8uXLQhmOtfE+/PBDcnV1pf3799PVq1dp165d1KdPH1q/fr1QhuNsuAMHDlBaWhrt3r2bANDevXt1zlsqpo2NjRQYGEhRUVFUWFhIWVlZpFKpKCEhweBr4qTGSGFhYbRo0SKdYwEBAbRs2TIr1ahnuXXrFgGg7OxsIiJqbm4mT09PWrNmjVDmwYMHpFQq6csvvyQiojt37pBEIqGdO3cKZSoqKkgsFtPBgweJiKi0tJQA0C+//CKUyc3NJQB0/vx5S1ya1dXV1dHAgQMpKyuLxo4dKyQ1HGPTSU1NpcjIyMee51ibRmxsLL366qs6x6ZPn05z5swhIo6zKTya1FgypgcOHCCxWEwVFRVCmR07dpBMJqPa2lqDroO7n4zQ0NCAgoICxMTE6ByPiYlBTk6OlWrVs9TW1gIAXFxcAABXr15FVVWVTkxlMhnGjh0rxLSgoAAPHz7UKaNSqRAYGCiUyc3NhVKpRHh4uFBm1KhRUCqVvea1eeuttxAbG4vx48frHOcYm86+ffswYsQIvPzyy/Dw8EBISAgyMzOF8xxr04iMjMTRo0dx8eJFAMBvv/2GU6dOYfLkyQA4zuZgyZjm5uYiMDAQKpVKKDNhwgTU19frdOV2Rq/a0NLU/vjjDzQ1NaFfv346x/v164eqqior1arnICKkpKQgMjISgYGBACDErb2YlpWVCWWkUin69u2rV6b196uqquDh4aH3nB4eHr3itdm5cycKCwvx66+/6p3jGJvOlStXsGXLFqSkpOCdd95Bfn4+EhMTIZPJEB8fz7E2kdTUVNTW1iIgIAB2dnZoampCRkYGZs2aBYD/ps3BkjGtqqrSe56+fftCKpUaHHdOakxAJBLp/ExEeseYvoSEBPz+++84deqU3rmuxPTRMu2V7w2vzfXr15GUlITDhw/DwcHhseU4xsZrbm7GiBEj8NFHHwEAQkJCcPbsWWzZsgXx8fFCOY61cb7//nts27YN3333HYYOHYri4mIkJydDpVJh3rx5QjmOs+lZKqamijt3PxnBzc0NdnZ2epnkrVu39LJOpmvx4sXYt28fjh07Bh8fH+G4p6cnAHQYU09PTzQ0NKCmpqbDMjdv3tR73urqapt/bQoKCnDr1i2EhobC3t4e9vb2yM7OxoYNG2Bvby9cP8fYeF5eXhgyZIjOscGDB6O8vBwA/z2bypIlS7Bs2TK88sorCAoKwty5c/H2229j9erVADjO5mDJmHp6euo9T01NDR4+fGhw3DmpMYJUKkVoaCiysrJ0jmdlZWHMmDFWqlX3RkRISEjAnj178PPPP8Pf31/nvL+/Pzw9PXVi2tDQgOzsbCGmoaGhkEgkOmU0Gg3OnDkjlBk9ejRqa2uRn58vlMnLy0Ntba3NvzbR0dEoKSlBcXGx8G/EiBGIi4tDcXEx+vfvzzE2kYiICL0lCS5evAg/Pz8A/PdsKvfu3YNYrPtxZWdnJ0zp5jibniVjOnr0aJw5cwYajUYoc/jwYchkMoSGhhpWcYOGFTM9rVO6v/nmGyotLaXk5GRSKBR07do1a1etW3rjjTdIqVTS8ePHSaPRCP/u3bsnlFmzZg0plUras2cPlZSU0KxZs9qdRujj40NHjhyhwsJCGjduXLvTCIcNG0a5ubmUm5tLQUFBNjs180nazn4i4hibSn5+Ptnb21NGRgZdunSJtm/fTo6OjrRt2zahDMfaePPmzSNvb29hSveePXvIzc2Nli5dKpThOBuurq6OioqKqKioiADQZ599RkVFRcKSJJaKaeuU7ujoaCosLKQjR46Qj48PT+m2lk2bNpGfnx9JpVJ65plnhOnJTB+Adv9t3bpVKNPc3Ezp6enk6elJMpmMnn32WSopKdF5nPv371NCQgK5uLiQXC6nKVOmUHl5uU6Z27dvU1xcHDk5OZGTkxPFxcVRTU2NBa6y+3k0qeEYm85PP/1EgYGBJJPJKCAggL7++mud8xxr42m1WkpKSiK1Wk0ODg7Uv39/SktLo/r6eqEMx9lwx44da/f9eN68eURk2ZiWlZVRbGwsyeVycnFxoYSEBHrw4IHB1yQiIjKsbYcxxhhjrPvhMTWMMcYYswmc1DDGGGPMJnBSwxhjjDGbwEkNY4wxxmwCJzWMMcYYswmc1DDGGGPMJnBSwxhjjDGbwEkNY4wxxmwCJzWMMcYYswmc1DDGGGPMJnBSwxhjjDGbwEkNY4wxxmzC/wNfIXr0ReB6qQAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x, y = stationaryDGP()\n",
    "spread = y - x\n",
    "df = pd.DataFrame({'x': x, 'y': y, 'close': spread}) # close price of portfolio is the spread\n",
    "df.plot(subplots=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                      y   R-squared:                       0.976\n",
      "Model:                            OLS   Adj. R-squared:                  0.976\n",
      "Method:                 Least Squares   F-statistic:                 4.073e+05\n",
      "Date:                Wed, 05 Jun 2024   Prob (F-statistic):               0.00\n",
      "Time:                        23:04:16   Log-Likelihood:                 746.19\n",
      "No. Observations:               10000   AIC:                            -1488.\n",
      "Df Residuals:                    9998   BIC:                            -1474.\n",
      "Df Model:                           1                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "==============================================================================\n",
      "                 coef    std err          t      P>|t|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "const         99.9939      0.017   5792.312      0.000      99.960     100.028\n",
      "x1             1.0012      0.002    638.198      0.000       0.998       1.004\n",
      "==============================================================================\n",
      "Omnibus:                       22.890   Durbin-Watson:                   0.212\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):               23.020\n",
      "Skew:                           0.112   Prob(JB):                     1.00e-05\n",
      "Kurtosis:                       3.071   Cond. No.                         85.3\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    }
   ],
   "source": [
    "# confirm the x, y pair is cointegrated, such that spread is stationary\n",
    "x_ols = sm.add_constant(x)\n",
    "ols = sm.OLS(y, x_ols).fit()\n",
    "print(ols.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ADF Statistic: -23.624303774542252\n",
      "p-value: 0.0\n",
      "Is the spread stationary? Yes\n"
     ]
    }
   ],
   "source": [
    "adfuller_test = adfuller(spread)\n",
    "print(f'ADF Statistic: {adfuller_test[0]}')\n",
    "print(f'p-value: {adfuller_test[1]}')\n",
    "print('Is the spread stationary? {}'.format('Yes' if adfuller_test[1] < 0.05 else 'No'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>x</th>\n",
       "      <th>y</th>\n",
       "      <th>close</th>\n",
       "      <th>feature_rolling_0</th>\n",
       "      <th>feature_rolling_1</th>\n",
       "      <th>feature_rolling_2</th>\n",
       "      <th>feature_rolling_3</th>\n",
       "      <th>feature_rolling_4</th>\n",
       "      <th>feature_rolling_5</th>\n",
       "      <th>feature_rolling_6</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9.952140</td>\n",
       "      <td>109.929877</td>\n",
       "      <td>99.977736</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9.953234</td>\n",
       "      <td>109.821062</td>\n",
       "      <td>99.867828</td>\n",
       "      <td>-0.054954</td>\n",
       "      <td>0.054954</td>\n",
       "      <td>0.054954</td>\n",
       "      <td>0.054954</td>\n",
       "      <td>0.054954</td>\n",
       "      <td>0.054954</td>\n",
       "      <td>0.054954</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9.970898</td>\n",
       "      <td>109.666858</td>\n",
       "      <td>99.695960</td>\n",
       "      <td>-0.077942</td>\n",
       "      <td>-0.015983</td>\n",
       "      <td>0.093925</td>\n",
       "      <td>0.093925</td>\n",
       "      <td>0.093925</td>\n",
       "      <td>0.093925</td>\n",
       "      <td>0.093925</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9.845759</td>\n",
       "      <td>109.517669</td>\n",
       "      <td>99.671910</td>\n",
       "      <td>0.052406</td>\n",
       "      <td>-0.095411</td>\n",
       "      <td>-0.033452</td>\n",
       "      <td>0.076457</td>\n",
       "      <td>0.076457</td>\n",
       "      <td>0.076457</td>\n",
       "      <td>0.076457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>9.835525</td>\n",
       "      <td>109.653331</td>\n",
       "      <td>99.817806</td>\n",
       "      <td>0.177882</td>\n",
       "      <td>0.007935</td>\n",
       "      <td>-0.139881</td>\n",
       "      <td>-0.077923</td>\n",
       "      <td>0.031986</td>\n",
       "      <td>0.031986</td>\n",
       "      <td>0.031986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9995</th>\n",
       "      <td>11.266501</td>\n",
       "      <td>111.250543</td>\n",
       "      <td>99.984042</td>\n",
       "      <td>-0.046866</td>\n",
       "      <td>-0.097044</td>\n",
       "      <td>-0.072946</td>\n",
       "      <td>0.090331</td>\n",
       "      <td>0.063302</td>\n",
       "      <td>-0.046692</td>\n",
       "      <td>0.109915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9996</th>\n",
       "      <td>11.264361</td>\n",
       "      <td>111.366934</td>\n",
       "      <td>100.102573</td>\n",
       "      <td>0.101266</td>\n",
       "      <td>-0.045425</td>\n",
       "      <td>-0.095603</td>\n",
       "      <td>-0.071504</td>\n",
       "      <td>0.091773</td>\n",
       "      <td>0.064743</td>\n",
       "      <td>-0.045251</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9997</th>\n",
       "      <td>11.329527</td>\n",
       "      <td>111.509216</td>\n",
       "      <td>100.179689</td>\n",
       "      <td>0.044838</td>\n",
       "      <td>0.086252</td>\n",
       "      <td>-0.060440</td>\n",
       "      <td>-0.110618</td>\n",
       "      <td>-0.086519</td>\n",
       "      <td>0.076758</td>\n",
       "      <td>0.049728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9998</th>\n",
       "      <td>11.295573</td>\n",
       "      <td>111.604869</td>\n",
       "      <td>100.309297</td>\n",
       "      <td>0.090529</td>\n",
       "      <td>0.038038</td>\n",
       "      <td>0.079452</td>\n",
       "      <td>-0.067240</td>\n",
       "      <td>-0.117418</td>\n",
       "      <td>-0.093319</td>\n",
       "      <td>0.069958</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9999</th>\n",
       "      <td>11.317716</td>\n",
       "      <td>111.744478</td>\n",
       "      <td>100.426762</td>\n",
       "      <td>0.077183</td>\n",
       "      <td>0.089325</td>\n",
       "      <td>0.036834</td>\n",
       "      <td>0.078248</td>\n",
       "      <td>-0.068444</td>\n",
       "      <td>-0.118622</td>\n",
       "      <td>-0.094523</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10000 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              x           y       close  feature_rolling_0  feature_rolling_1  \\\n",
       "0      9.952140  109.929877   99.977736           0.000000           0.000000   \n",
       "1      9.953234  109.821062   99.867828          -0.054954           0.054954   \n",
       "2      9.970898  109.666858   99.695960          -0.077942          -0.015983   \n",
       "3      9.845759  109.517669   99.671910           0.052406          -0.095411   \n",
       "4      9.835525  109.653331   99.817806           0.177882           0.007935   \n",
       "...         ...         ...         ...                ...                ...   \n",
       "9995  11.266501  111.250543   99.984042          -0.046866          -0.097044   \n",
       "9996  11.264361  111.366934  100.102573           0.101266          -0.045425   \n",
       "9997  11.329527  111.509216  100.179689           0.044838           0.086252   \n",
       "9998  11.295573  111.604869  100.309297           0.090529           0.038038   \n",
       "9999  11.317716  111.744478  100.426762           0.077183           0.089325   \n",
       "\n",
       "      feature_rolling_2  feature_rolling_3  feature_rolling_4  \\\n",
       "0              0.000000           0.000000           0.000000   \n",
       "1              0.054954           0.054954           0.054954   \n",
       "2              0.093925           0.093925           0.093925   \n",
       "3             -0.033452           0.076457           0.076457   \n",
       "4             -0.139881          -0.077923           0.031986   \n",
       "...                 ...                ...                ...   \n",
       "9995          -0.072946           0.090331           0.063302   \n",
       "9996          -0.095603          -0.071504           0.091773   \n",
       "9997          -0.060440          -0.110618          -0.086519   \n",
       "9998           0.079452          -0.067240          -0.117418   \n",
       "9999           0.036834           0.078248          -0.068444   \n",
       "\n",
       "      feature_rolling_5  feature_rolling_6  \n",
       "0              0.000000           0.000000  \n",
       "1              0.054954           0.054954  \n",
       "2              0.093925           0.093925  \n",
       "3              0.076457           0.076457  \n",
       "4              0.031986           0.031986  \n",
       "...                 ...                ...  \n",
       "9995          -0.046692           0.109915  \n",
       "9996           0.064743          -0.045251  \n",
       "9997           0.076758           0.049728  \n",
       "9998          -0.093319           0.069958  \n",
       "9999          -0.118622          -0.094523  \n",
       "\n",
       "[10000 rows x 10 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# build features\n",
    "df_train = df.copy()\n",
    "\n",
    "feature_config = {\n",
    "    \"window\": 7, # 7 days window, pad with first value when window is not available\n",
    "    \"mode\": \"diff\", # difference between current and previous close price\n",
    "    \"subtract_mean\": True, # subtract mean from the window\n",
    "}\n",
    "\n",
    "fe = FeatureEngineering(df_train, main_feature_col=\"close\")\n",
    "fe.rolling_feature(**feature_config)\n",
    "df_train = fe.df.copy()\n",
    "df_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Market Return :  0.45%   |   Portfolio Return : -60.52%   |   Position Changes : 6665   |   Episode Length : 10000   |   \n"
     ]
    }
   ],
   "source": [
    "# Define the environment\n",
    "env_action_space = [-1, 0, 1] # Positions : [-1=SHORT, 0=OUT, 1=LONG]\n",
    "# env_action_space = [i/10 for i in range(-10, 11)] # Positions : [-1, -0.9, -0.8, ..., 0, 0.1, 0.2, ..., 1]\n",
    "\n",
    "env = gym.make(\"TradingEnv\",\n",
    "        name= \"stationaryDGP\",\n",
    "        df = df_train, # Your dataset with your custom features\n",
    "        positions = env_action_space, # -1 (=SHORT), 0(=OUT), +1 (=LONG)\n",
    "        trading_fees = 0.01/100, # 0.01% per stock buy / sell (Binance fees)\n",
    "        borrow_interest_rate= 0.0003/100, # 0.0003% per timestep (one timestep = 1h here)\n",
    "        # max_episode_duration=180\n",
    "    )\n",
    "env.unwrapped.add_metric('Position Changes', lambda history : np.sum(np.diff(history['position']) != 0) )\n",
    "env.unwrapped.add_metric('Episode Length', lambda history : len(history['position']) )\n",
    "\n",
    "# Run an episode until it ends :\n",
    "done, truncated = False, False\n",
    "observation, info = env.reset()\n",
    "print(observation)\n",
    "\n",
    "while not done and not truncated:\n",
    "    # Pick a position by its index in your position list (=[-1, 0, 1])....usually something like : position_index = your_policy(observation)\n",
    "    position_index = env.action_space.sample() # At every timestep, pick a random position index from your position list (=[-1, 0, 1])\n",
    "    observation, reward, done, truncated, info = env.step(position_index)\n",
    "\n",
    "n_actions = env.action_space.n\n",
    "n_observations = len(env.reset()[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Market Return :  0.45%   |   Portfolio Return : -44.63%   |   Position Changes : 5721   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -40.92%   |   Position Changes : 4918   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -31.06%   |   Position Changes : 4837   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -36.59%   |   Position Changes : 4733   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -33.96%   |   Position Changes : 4705   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -29.39%   |   Position Changes : 4585   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -28.24%   |   Position Changes : 4738   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -37.22%   |   Position Changes : 4608   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -16.35%   |   Position Changes : 4518   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -26.47%   |   Position Changes : 4458   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -29.19%   |   Position Changes : 4418   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -19.66%   |   Position Changes : 4253   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -22.48%   |   Position Changes : 4343   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -10.36%   |   Position Changes : 4316   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -1.43%   |   Position Changes : 4281   |   Episode Length : 10000   |   \n"
     ]
    }
   ],
   "source": [
    "# Train a DQN agent using training data from a stationary data generating process\n",
    "agent = DQNAgent(n_observations=n_observations, n_actions=n_actions)\n",
    "history_metrics, _ = agent.train(env, num_episodes=15) # Beware of overfitting when num_episodes is too high"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "all_metrics = pd.DataFrame(history_metrics)\n",
    "perc2float = lambda x: float(x.strip('%'))/100\n",
    "\n",
    "all_metrics[\"Market Return\"] = all_metrics[\"Market Return\"].apply(perc2float)\n",
    "all_metrics[\"Portfolio Return\"] = all_metrics[\"Portfolio Return\"].apply(perc2float)\n",
    "\n",
    "all_metrics.plot(y=[\"episodic_reward\"], figsize=(15, 5))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Market Return :  0.22%   |   Portfolio Return : -19.20%   |   Position Changes : 2234   |   Episode Length : 10000   |   \n",
      "Market Return : -0.13%   |   Portfolio Return : -13.09%   |   Position Changes : 2320   |   Episode Length : 10000   |   \n",
      "Market Return :  0.03%   |   Portfolio Return : -20.73%   |   Position Changes : 2252   |   Episode Length : 10000   |   \n",
      "Market Return : -0.07%   |   Portfolio Return : -21.50%   |   Position Changes : 2277   |   Episode Length : 10000   |   \n",
      "Market Return :  0.48%   |   Portfolio Return : -22.00%   |   Position Changes : 2185   |   Episode Length : 10000   |   \n",
      "Market Return :  0.05%   |   Portfolio Return : -24.78%   |   Position Changes : 2234   |   Episode Length : 10000   |   \n",
      "Market Return : -0.01%   |   Portfolio Return : -16.86%   |   Position Changes : 2252   |   Episode Length : 10000   |   \n",
      "Market Return : -0.01%   |   Portfolio Return : -20.26%   |   Position Changes : 2220   |   Episode Length : 10000   |   \n",
      "Market Return : -0.02%   |   Portfolio Return : -15.39%   |   Position Changes : 2258   |   Episode Length : 10000   |   \n",
      "Market Return : -0.02%   |   Portfolio Return : -20.15%   |   Position Changes : 2326   |   Episode Length : 10000   |   \n",
      "Average Portfolio Return: -0.19396\n",
      "Std Portfolio Return: 0.0324920051705031\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the agent on a new dataset from the same data generating process\n",
    "returns = []\n",
    "for i in range(10):\n",
    "    x, y = stationaryDGP()\n",
    "    spread = y - x\n",
    "    _df = pd.DataFrame({'x': x, 'y': y, 'close': spread})\n",
    "    fe = FeatureEngineering(_df, main_feature_col=\"close\")\n",
    "    fe.rolling_feature(**feature_config)\n",
    "    _df = fe.df.copy()\n",
    "    \n",
    "    env_test = gym.make(\"TradingEnv\",\n",
    "            name= \"stationaryDGP\",\n",
    "            df = _df, # Your dataset with your custom features\n",
    "            positions = env_action_space, # -1 (=SHORT), 0(=OUT), +1 (=LONG)\n",
    "            trading_fees = 0.01/100, # 0.01% per stock buy / sell (Binance fees)\n",
    "            borrow_interest_rate= 0.0003/100, # 0.0003% per timestep (one timestep = 1h here)\n",
    "        )\n",
    "    env_test.unwrapped.add_metric('Position Changes', lambda history : np.sum(np.diff(history['position']) != 0) )\n",
    "    env_test.unwrapped.add_metric('Episode Length', lambda history : len(history['position']) )\n",
    "\n",
    "    env_test = agent.eval(env_test)\n",
    "    returns.append(perc2float(env_test.unwrapped.get_metrics()[\"Portfolio Return\"]))\n",
    "\n",
    "print(f\"Average Portfolio Return: {np.mean(returns)}\")\n",
    "print(f\"Std Portfolio Return: {np.std(returns)}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from gym_trading_env.renderer import Renderer\n",
    "# renderer = Renderer(render_logs_dir=\"render_logs\")\n",
    "# renderer.run()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>x</th>\n",
       "      <th>y</th>\n",
       "      <th>close</th>\n",
       "      <th>feature_rolling_0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9.952140</td>\n",
       "      <td>109.929877</td>\n",
       "      <td>99.977736</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9.953234</td>\n",
       "      <td>109.821062</td>\n",
       "      <td>99.867828</td>\n",
       "      <td>-0.109909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9.970898</td>\n",
       "      <td>109.666858</td>\n",
       "      <td>99.695960</td>\n",
       "      <td>-0.171867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9.845759</td>\n",
       "      <td>109.517669</td>\n",
       "      <td>99.671910</td>\n",
       "      <td>-0.024051</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>9.835525</td>\n",
       "      <td>109.653331</td>\n",
       "      <td>99.817806</td>\n",
       "      <td>0.145896</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9995</th>\n",
       "      <td>11.266501</td>\n",
       "      <td>111.250543</td>\n",
       "      <td>99.984042</td>\n",
       "      <td>-0.028161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9996</th>\n",
       "      <td>11.264361</td>\n",
       "      <td>111.366934</td>\n",
       "      <td>100.102573</td>\n",
       "      <td>0.118530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9997</th>\n",
       "      <td>11.329527</td>\n",
       "      <td>111.509216</td>\n",
       "      <td>100.179689</td>\n",
       "      <td>0.077117</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9998</th>\n",
       "      <td>11.295573</td>\n",
       "      <td>111.604869</td>\n",
       "      <td>100.309297</td>\n",
       "      <td>0.129607</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9999</th>\n",
       "      <td>11.317716</td>\n",
       "      <td>111.744478</td>\n",
       "      <td>100.426762</td>\n",
       "      <td>0.117465</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10000 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              x           y       close  feature_rolling_0\n",
       "0      9.952140  109.929877   99.977736           0.000000\n",
       "1      9.953234  109.821062   99.867828          -0.109909\n",
       "2      9.970898  109.666858   99.695960          -0.171867\n",
       "3      9.845759  109.517669   99.671910          -0.024051\n",
       "4      9.835525  109.653331   99.817806           0.145896\n",
       "...         ...         ...         ...                ...\n",
       "9995  11.266501  111.250543   99.984042          -0.028161\n",
       "9996  11.264361  111.366934  100.102573           0.118530\n",
       "9997  11.329527  111.509216  100.179689           0.077117\n",
       "9998  11.295573  111.604869  100.309297           0.129607\n",
       "9999  11.317716  111.744478  100.426762           0.117465\n",
       "\n",
       "[10000 rows x 4 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# build features\n",
    "df_train = df.copy()\n",
    "\n",
    "feature_config = {\n",
    "    \"window\": 1, # 7 days window, pad with first value when window is not available\n",
    "    \"mode\": \"diff\" # difference between current and previous close price\n",
    "}\n",
    "\n",
    "fe = FeatureEngineering(df_train, main_feature_col=\"close\")\n",
    "fe.rolling_feature(**feature_config)\n",
    "df_train = fe.df.copy()\n",
    "df_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0. -1. -1.]\n",
      "Market Return :  0.45%   |   Portfolio Return : -0.45%   |   Position Changes : 0   |   Episode Length : 10000   |   \n"
     ]
    }
   ],
   "source": [
    "# Define the environment\n",
    "env_action_space = [-1, 0, 1] # Positions : [-1=SHORT, 0=OUT, 1=LONG]\n",
    "# env_action_space = [i/10 for i in range(-10, 11)] # Positions : [-1, -0.9, -0.8, ..., 0, 0.1, 0.2, ..., 1]\n",
    "\n",
    "env = gym.make(\"TradingEnv\",\n",
    "        name= \"stationaryDGP\",\n",
    "        df = df_train, # Your dataset with your custom features\n",
    "        positions = env_action_space, # -1 (=SHORT), 0(=OUT), +1 (=LONG)\n",
    "        trading_fees = 0.01/100, # 0.01% per stock buy / sell (Binance fees)\n",
    "        borrow_interest_rate= 0.0003/100, # 0.0003% per timestep (one timestep = 1h here)\n",
    "        # max_episode_duration=180\n",
    "    )\n",
    "env.unwrapped.add_metric('Position Changes', lambda history : np.sum(np.diff(history['position']) != 0) )\n",
    "env.unwrapped.add_metric('Episode Length', lambda history : len(history['position']) )\n",
    "\n",
    "# Run an episode until it ends :\n",
    "done, truncated = False, False\n",
    "observation, info = env.reset()\n",
    "print(observation)\n",
    "while not done and not truncated:\n",
    "    # Pick a position by its index in your position list (=[-1, 0, 1])....usually something like : position_index = your_policy(observation)\n",
    "    position_index = env.action_space.sample() # At every timestep, pick a random position index from your position list (=[-1, 0, 1])\n",
    "    observation, reward, done, truncated, info = env.step(0)\n",
    "\n",
    "n_actions = env.action_space.n\n",
    "n_observations = len(env.reset()[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Market Return :  0.45%   |   Portfolio Return : -59.12%   |   Position Changes : 6706   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -54.75%   |   Position Changes : 6705   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -54.71%   |   Position Changes : 5725   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -40.27%   |   Position Changes : 5670   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -14.17%   |   Position Changes : 4437   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : -24.24%   |   Position Changes : 4442   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return :  2.74%   |   Position Changes : 2647   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : 11.85%   |   Position Changes : 2516   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : 22.87%   |   Position Changes : 1876   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : 19.50%   |   Position Changes : 1838   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : 30.66%   |   Position Changes : 1584   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : 31.19%   |   Position Changes : 1585   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : 35.81%   |   Position Changes : 1481   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : 32.50%   |   Position Changes : 1492   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : 38.03%   |   Position Changes : 1467   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : 33.16%   |   Position Changes : 1491   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : 43.37%   |   Position Changes : 1389   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : 41.35%   |   Position Changes : 1394   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : 54.15%   |   Position Changes : 1382   |   Episode Length : 10000   |   \n",
      "Market Return :  0.45%   |   Portfolio Return : 55.46%   |   Position Changes : 1366   |   Episode Length : 10000   |   \n"
     ]
    }
   ],
   "source": [
    "agent = XGBoostAgent(n_observations, n_actions)\n",
    "history_metrics, _ = agent.train(env, num_episodes=20, replay_every=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "all_metrics = pd.DataFrame(history_metrics)\n",
    "perc2float = lambda x: float(x.strip('%'))/100\n",
    "\n",
    "all_metrics[\"Market Return\"] = all_metrics[\"Market Return\"].apply(perc2float)\n",
    "all_metrics[\"Portfolio Return\"] = all_metrics[\"Portfolio Return\"].apply(perc2float)\n",
    "\n",
    "all_metrics.plot(y=[\"episodic_reward\"], figsize=(15, 5))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Market Return :  0.32%   |   Portfolio Return : 37.63%   |   Position Changes : 1343   |   Episode Length : 10000   |   \n",
      "Market Return : -0.29%   |   Portfolio Return : 40.41%   |   Position Changes : 1341   |   Episode Length : 10000   |   \n",
      "Market Return : -0.11%   |   Portfolio Return : 46.46%   |   Position Changes : 1310   |   Episode Length : 10000   |   \n",
      "Market Return :  0.35%   |   Portfolio Return : 39.26%   |   Position Changes : 1362   |   Episode Length : 10000   |   \n",
      "Market Return : -0.12%   |   Portfolio Return : 43.92%   |   Position Changes : 1314   |   Episode Length : 10000   |   \n",
      "Market Return : -0.09%   |   Portfolio Return : 49.58%   |   Position Changes : 1374   |   Episode Length : 10000   |   \n",
      "Market Return : -0.17%   |   Portfolio Return : 58.70%   |   Position Changes : 1373   |   Episode Length : 10000   |   \n",
      "Market Return :  0.07%   |   Portfolio Return : 44.11%   |   Position Changes : 1314   |   Episode Length : 10000   |   \n",
      "Market Return :  0.13%   |   Portfolio Return : 49.63%   |   Position Changes : 1337   |   Episode Length : 10000   |   \n",
      "Market Return : -0.23%   |   Portfolio Return : 40.05%   |   Position Changes : 1265   |   Episode Length : 10000   |   \n",
      "Average Portfolio Return: 0.44975\n",
      "Std Portfolio Return: 0.06058236129435697\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the agent on a new dataset from the same data generating process\n",
    "returns = []\n",
    "for i in range(10):\n",
    "    x, y = stationaryDGP()\n",
    "    spread = y - x\n",
    "    _df = pd.DataFrame({'x': x, 'y': y, 'close': spread})\n",
    "    fe = FeatureEngineering(_df, main_feature_col=\"close\")\n",
    "    fe.rolling_feature(**feature_config)\n",
    "    _df = fe.df.copy()\n",
    "\n",
    "    env_test = gym.make(\"TradingEnv\",\n",
    "            name= \"stationaryDGP\",\n",
    "            df = _df, # Your dataset with your custom features\n",
    "            positions = env_action_space, # -1 (=SHORT), 0(=OUT), +1 (=LONG)\n",
    "            trading_fees = 0.01/100, # 0.01% per stock buy / sell (Binance fees)\n",
    "            borrow_interest_rate= 0.0003/100, # 0.0003% per timestep (one timestep = 1h here)\n",
    "        )\n",
    "    env_test.unwrapped.add_metric('Position Changes', lambda history : np.sum(np.diff(history['position']) != 0) )\n",
    "    env_test.unwrapped.add_metric('Episode Length', lambda history : len(history['position']) )\n",
    "\n",
    "    env_test = agent.eval(env_test)\n",
    "    returns.append(perc2float(env_test.unwrapped.get_metrics()[\"Portfolio Return\"]))\n",
    "\n",
    "print(f\"Average Portfolio Return: {np.mean(returns)}\")\n",
    "print(f\"Std Portfolio Return: {np.std(returns)}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "env_full_py3_9",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}