Data Analysis and Pre-Processing(Alpha 191).ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "6c5bb934",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-04-27T03:06:31.184586Z",
     "start_time": "2023-04-27T03:06:28.617889Z"
    }
   },
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "import pandas as pd\n",
    "#显示Dateframe所有列(参数设置为None代表显示所有行，也可以自行设置数字)\n",
    "pd.set_option('display.max_columns',None)\n",
    "#显示Dateframe所有行\n",
    "pd.set_option('display.max_rows',None)\n",
    "#设置Dataframe数据的显示长度，默认为50\n",
    "pd.set_option('max_colwidth',200)\n",
    "#不显示科学计数法，保留5位小数\n",
    "pd.set_option('display.float_format', lambda x: '%.5f' % x)\n",
    "import mplfinance as mpf\n",
    "import numpy as np\n",
    "import os\n",
    "from tqdm import tqdm\n",
    "from IPython.display import clear_output\n",
    "from joblib import *\n",
    "from alpha_factors.GTJA_Alpha191_analysis import GET_alpha"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "e6e5eda4",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-04-27T03:06:31.854024Z",
     "start_time": "2023-04-27T03:06:31.777974Z"
    },
    "code_folding": [
     3
    ],
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((6558, 10),\n",
       "                   date tic underlaying_symbols       open       high  \\\n",
       " 0  2022-08-24 21:00:00  RB         SHFE.rb2210 4082.00000 4089.00000   \n",
       " 1  2022-08-24 21:05:00  RB         SHFE.rb2210 4089.00000 4100.00000   \n",
       " 2  2022-08-24 21:10:00  RB         SHFE.rb2210 4086.00000 4095.00000   \n",
       " 3  2022-08-24 21:15:00  RB         SHFE.rb2210 4085.00000 4090.00000   \n",
       " 4  2022-08-24 21:20:00  RB         SHFE.rb2210 4084.00000 4091.00000   \n",
       " \n",
       "          low      close      volume       open_oi      close_oi  \n",
       " 0 4081.00000 4086.00000 45466.00000 1289350.00000 1288077.00000  \n",
       " 1 4083.00000 4086.00000 65069.00000 1288077.00000 1286786.00000  \n",
       " 2 4082.00000 4085.00000 23530.00000 1286786.00000 1285704.00000  \n",
       " 3 4079.00000 4084.00000 18618.00000 1285704.00000 1284464.00000  \n",
       " 4 4082.00000 4091.00000 16848.00000 1284464.00000 1283701.00000  ,\n",
       " date                    object\n",
       " tic                     object\n",
       " underlaying_symbols     object\n",
       " open                   float64\n",
       " high                   float64\n",
       " low                    float64\n",
       " close                  float64\n",
       " volume                 float64\n",
       " open_oi                float64\n",
       " close_oi               float64\n",
       " dtype: object)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "path = 'C:\\\\Users\\\\budin\\\\Desktop\\\\强化学习\\\\螺纹钢\\\\5分钟 - 副本\\\\数据\\\\'\n",
    "dir_list = os.listdir(path)\n",
    "train_df = pd.DataFrame()\n",
    "for i in range(len(dir_list)):\n",
    "    if 'train' in dir_list[i]:\n",
    "        tmp_df = pd.read_csv(path+dir_list[i])\n",
    "        train_df = train_df.append(tmp_df)\n",
    "    else:\n",
    "        pass\n",
    "train_df = train_df.drop(columns='Unnamed: 0').rename(columns={'symbol':'underlaying_symbols', 'datetime': 'date'})\n",
    "train_df['tic'] = ['RB']*len(train_df)\n",
    "train_df = train_df[['date', 'tic', 'underlaying_symbols', 'open', 'high', 'low', 'close', 'volume', 'open_oi', 'close_oi']]\n",
    "tickers = ['RB']\n",
    "\n",
    "train_df.shape, train_df.head(5), train_df.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8f7a7880",
   "metadata": {},
   "source": [
    "# Reduce Contract Change Gap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "9e366329",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-04-27T03:06:34.171542Z",
     "start_time": "2023-04-27T03:06:34.153529Z"
    },
    "code_folding": [
     2
    ],
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[Int64Index([  0,   1,   2,   3,   4,   5,   6,   7,   8,   9,\n",
       "              ...\n",
       "              210, 211, 212, 213, 214, 215, 216, 217, 218, 219],\n",
       "             dtype='int64', length=220)],\n",
       " [Int64Index([ 220,  221,  222,  223,  224,  225,  226,  227,  228,  229,\n",
       "              ...\n",
       "              4127, 4128, 4129, 4130, 4131, 4132, 4133, 4134, 4135, 4136],\n",
       "             dtype='int64', length=3917)],\n",
       " [Int64Index([4137, 4138, 4139, 4140, 4141, 4142, 4143, 4144, 4145, 4146,\n",
       "              ...\n",
       "              6548, 6549, 6550, 6551, 6552, 6553, 6554, 6555, 6556, 6557],\n",
       "             dtype='int64', length=2421)]]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "contracts = train_df['underlaying_symbols'].unique()\n",
    "indexes = []\n",
    "for i in range(len(contracts)):\n",
    "    tmp_df = train_df[train_df['underlaying_symbols']==contracts[i]]\n",
    "    tmp_index = tmp_df.index\n",
    "    indexes.append([tmp_index])\n",
    "    \n",
    "indexes"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e6dfda08",
   "metadata": {},
   "source": [
    "## Inspect Individual Price Gaps "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "3c631a72",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-04-27T03:06:36.360000Z",
     "start_time": "2023-04-27T03:06:35.972746Z"
    },
    "code_folding": [
     0
    ],
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x575 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x575 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(1, len(contracts)):\n",
    "    last_index_end = indexes[i-1][0][-5]\n",
    "    this_index_start = indexes[i][0][5]\n",
    "    tmp_df = train_df.iloc[last_index_end:this_index_start].set_index('date')\n",
    "    tmp_df.index = pd.to_datetime(tmp_df.index)\n",
    "    mpf.plot(tmp_df, type='candle')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a4c252f2",
   "metadata": {},
   "source": [
    "## Adjust Individual Price Gaps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "885f1510",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-04-27T03:06:37.478733Z",
     "start_time": "2023-04-27T03:06:37.437705Z"
    },
    "code_folding": [
     1
    ]
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-209.0\n",
      "-235.0\n"
     ]
    },
    {
     "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>date</th>\n",
       "      <th>tic</th>\n",
       "      <th>underlaying_symbols</th>\n",
       "      <th>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>volume</th>\n",
       "      <th>open_oi</th>\n",
       "      <th>close_oi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2022-08-24 21:00:00</td>\n",
       "      <td>RB</td>\n",
       "      <td>SHFE.rb2210</td>\n",
       "      <td>4082.00000</td>\n",
       "      <td>4089.00000</td>\n",
       "      <td>4081.00000</td>\n",
       "      <td>4086.00000</td>\n",
       "      <td>45466.00000</td>\n",
       "      <td>1289350.00000</td>\n",
       "      <td>1288077.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2022-08-24 21:05:00</td>\n",
       "      <td>RB</td>\n",
       "      <td>SHFE.rb2210</td>\n",
       "      <td>4089.00000</td>\n",
       "      <td>4100.00000</td>\n",
       "      <td>4083.00000</td>\n",
       "      <td>4086.00000</td>\n",
       "      <td>65069.00000</td>\n",
       "      <td>1288077.00000</td>\n",
       "      <td>1286786.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2022-08-24 21:10:00</td>\n",
       "      <td>RB</td>\n",
       "      <td>SHFE.rb2210</td>\n",
       "      <td>4086.00000</td>\n",
       "      <td>4095.00000</td>\n",
       "      <td>4082.00000</td>\n",
       "      <td>4085.00000</td>\n",
       "      <td>23530.00000</td>\n",
       "      <td>1286786.00000</td>\n",
       "      <td>1285704.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2022-08-24 21:15:00</td>\n",
       "      <td>RB</td>\n",
       "      <td>SHFE.rb2210</td>\n",
       "      <td>4085.00000</td>\n",
       "      <td>4090.00000</td>\n",
       "      <td>4079.00000</td>\n",
       "      <td>4084.00000</td>\n",
       "      <td>18618.00000</td>\n",
       "      <td>1285704.00000</td>\n",
       "      <td>1284464.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2022-08-24 21:20:00</td>\n",
       "      <td>RB</td>\n",
       "      <td>SHFE.rb2210</td>\n",
       "      <td>4084.00000</td>\n",
       "      <td>4091.00000</td>\n",
       "      <td>4082.00000</td>\n",
       "      <td>4091.00000</td>\n",
       "      <td>16848.00000</td>\n",
       "      <td>1284464.00000</td>\n",
       "      <td>1283701.00000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  date tic underlaying_symbols       open       high  \\\n",
       "0  2022-08-24 21:00:00  RB         SHFE.rb2210 4082.00000 4089.00000   \n",
       "1  2022-08-24 21:05:00  RB         SHFE.rb2210 4089.00000 4100.00000   \n",
       "2  2022-08-24 21:10:00  RB         SHFE.rb2210 4086.00000 4095.00000   \n",
       "3  2022-08-24 21:15:00  RB         SHFE.rb2210 4085.00000 4090.00000   \n",
       "4  2022-08-24 21:20:00  RB         SHFE.rb2210 4084.00000 4091.00000   \n",
       "\n",
       "         low      close      volume       open_oi      close_oi  \n",
       "0 4081.00000 4086.00000 45466.00000 1289350.00000 1288077.00000  \n",
       "1 4083.00000 4086.00000 65069.00000 1288077.00000 1286786.00000  \n",
       "2 4082.00000 4085.00000 23530.00000 1286786.00000 1285704.00000  \n",
       "3 4079.00000 4084.00000 18618.00000 1285704.00000 1284464.00000  \n",
       "4 4082.00000 4091.00000 16848.00000 1284464.00000 1283701.00000  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "price_adjs = []\n",
    "for i in range(1, len(contracts)):\n",
    "    tmp_index = indexes[i][0]\n",
    "    last_price = train_df['close'].iloc[indexes[i-1][0][-1]]\n",
    "    now_price = train_df['close'].iloc[indexes[i][0][0]]\n",
    "    price_adj = now_price - last_price\n",
    "    print(price_adj)\n",
    "    train_df['close'].iloc[tmp_index] = train_df['close'].iloc[tmp_index] - price_adj\n",
    "    train_df['high'].iloc[tmp_index] = train_df['high'].iloc[tmp_index] - price_adj\n",
    "    train_df['low'].iloc[tmp_index] = train_df['low'].iloc[tmp_index] - price_adj\n",
    "    train_df['open'].iloc[tmp_index] = train_df['open'].iloc[tmp_index] - price_adj\n",
    "    \n",
    "train_df.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b747e7a0",
   "metadata": {},
   "source": [
    "## Varify Individual Price Gaps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "ea67f221",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-04-27T03:06:38.692528Z",
     "start_time": "2023-04-27T03:06:38.365316Z"
    },
    "code_folding": [
     0
    ],
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAApYAAAH/CAYAAADt4Z0FAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAA9hAAAPYQGoP6dpAABak0lEQVR4nO3dfVwVZf7/8fcBDiogoiLmDQFKeIumomXelUG1uWq66Lq7rneZioqlqdmWm2V3W2qGoklquVG7amjmtr9MWy13a1O/upaaaIh3qJCaICACh/n9QZwkreQwOgd8PR8PHugwN9fMh3POm5m5rrEZhmEIAAAAqCQPqxsAAACA6oFgCQAAAFMQLAEAAGAKgiUAAABMQbAEAACAKQiWAAAAMAXBEgAAAKbwsroBVUVJSYlOnz4tSfLx8ZHNZrO4RQAAANeWYRjKz8+XJAUGBsrD4+fPSRIsr9Lp06fVsGFDq5sBAABgiczMTAUFBf3sPFwKBwAAgCk4Y3mVfHx8nP/+764D5f5/o7PZpLDGAUo/cU48INQa1MB61MB61MB61MB6ZtcgPz9ft3eIkKSryj4Ey6t06T2VPj4+8vHxtbA17sVmk3x9feXjU8QbiUWogfWogfWogfWogfWuZQ2upn8Jl8IBAABgCoIlAAAATEGwBAAAgCkIlgAAADAFwRIAAACmIFgCAADAFARLAAAAmIJgCQAAAFMQLAEAAGAKgiUAAABMQbAEAACAKQiWAAAAMAXBEgAAAKYgWAIAAMAUBEsAAACYgmAJwC1kZZ7Sq/NeVFbmKaubAgBwEcESgFvIysrUglf+oqysTKubAgBwEcESAAAApiBYAgAAwBQESwAAAJiCYAkAAABTECyBaiAr85RmzZpFj2oAgKUIlkA1kJWVqaeffpoe1QAASxEsAQAAYAqCJQAAAExBsAQAAIApCJYAAAAwBcESAAAApiBYAgAAwBRuFywdDoeSkpIUExOjdu3aqV+/flq3bl25eTZv3qzY2FhFRkaqR48eevbZZ5Wbm1tunjlz5qhFixaXfSUlJZWb74033lB0dLQiIyPVv39/bdq06ZrvIwAAQHXkZXUDfmzevHlasWKFJk2apMjISH3yySeaPn26PDw81LdvX23cuFHx8fHq0qWL5s+fr+LiYi1evFjDhw/XypUr5eVVuktff/21unbtqkceeaTc+hs1auT899KlSzVv3jxNmDBBbdu2VUpKiiZNmqQVK1aoc+fO13O3AQAAqjy3CpZ5eXlKTk7W8OHDNWbMGElS165dtXfvXiUnJ6tv375asGCBwsPDtXTpUnl7e0uSoqKiFB0drTVr1mjw4MGSpP379+t3v/udbr311ituq6CgQEuWLNGIESM0YcIESVLPnj01ZMgQJSYm6s0337zm+wsAAFCduNWl8Bo1amjlypUaOXJkuel2u12FhYWSpEOHDql79+7OUClJ9evXV7NmzbR582ZJ0unTp3X69Gm1bNnyJ7e1e/du5eTk6J577nFOs9lsiomJ0bZt21RQUGDmrgEAAFR7bnXG0svLyxkGDcPQ6dOntWbNGn322WeaPXu2JKlu3brKyMgot1xRUZFOnjzpDJ/79u2TJH388cd67rnnlJWVpVtuuUWTJ09Wr169JElpaWmSpNDQ0HLrCgkJkcPh0NGjRxUREXHFdtpspV8oVXYsOCbWubQGVtYh/VCa8vJyf3nGKziUdsD53ZV98PX1U1iz5i5t2wy8DqxHDaxHDaxndg0quh63CpaXWr9+vaZNmyZJ6tWrl+6//35J0sCBA/Xaa68pKSlJsbGxKigo0Pz585WbmysfHx9JpZfBJens2bN69tlnVVhYqOTkZI0bN05JSUnq0aOHzp8/L0ny8/Mrt11fX19Juqwz0KXCGgc454N08uRJzZo1X2PHjlWjxo1+eQGY7lxmbUlSkwa11axJgCVtOHjwoKJ7RlV6PVPix7i87IEDB3TLLbdUug2VEdY4wNLtgxq4A2pgPbNqkJdnr9D8bhss27dvr+TkZKWnpyshIUFDhgzRu+++q/j4eDkcDiUkJGju3Lmy2+0aNGiQoqOjdfDgQUlSnz591KJFC/Xo0UMeHqVX+7t3767+/fsrISFBPXr0UElJyRW3axiGJDmXu5L0E+fk41Nk8h5XXXv3HNDTTz+tTl3vUpuSWlY354aU8e155/eAjHOWtOHrtBOSpBV//atatmx1Xbe9f//XGj5smL5OOyFPnwbXddtlbLbSN/L0E+f0/dsIrjNqYD1qYD2za5Cfn1eh+d02WIaEhCgkJESdO3dWcHCwRowYoQ0bNqhfv36aOnWq4uPjdezYMQUFBcnf319Dhw5VQECAJKlJkyZq0qRJufXZ7XZ169ZNK1eulCT5+/tLKu0wVKdOHed8+fn5kqTatWv/ZNsMQ7xgLlF2LDgu1nGHGpRtt2XLVurYsaNlbbD6d9Ad2nCjowbWowbWM6sGFV2HW3XeOXPmjNauXaszZ86Umx4ZGSlJOnXqlLZt26atW7eqRo0aCg8Pl7+/v4qLi5WamqrWrVtLkrZs2aKPPvrosvVfvHjRGT7DwsIkSUeOHCk3z5EjR+Tt7a3g4GCzdw8AAKBac6tgmZ+frxkzZmj16tXlpm/dulWS1KJFC3344YeaOXOmiop+uBSdkpKinJwcxcTESJL++c9/6vHHH1d2dna5dW/ZskVdunSRJHXo0EE+Pj7asGGDcx7DMLRx40Z16dKlXK9zAAAA/DK3uhQeHBysBx54QImJifLw8FBkZKT27NmjxYsXq3v37urZs6caNWqkVatWacaMGYqNjVVqaqrmzJmjPn36KCqqtOPA6NGjtWHDBo0ZM0ZjxoyRw+HQ66+/rvz8fE2aNEmSVKtWLY0aNUqJiYmy2+3q0KGDUlJStHfvXq1YscLKwwAAAFAluVWwlKTZs2crNDRUKSkpWrBggRo0aKBhw4Zp/PjxstlsioiI0JIlSzR37lyNGzdOgYGBGjdunMaOHetcR0REhJKTkzV//nz96U9/UmFhoTp37qznnntON998s3O+iRMnytPTU6tWrdLy5csVHh6uRYsWqVOnTlbsOgAAQJXmdsHS29tbcXFxiouL+8l5unXrpm7duv3seiIjI7Vs2bKfncdms2n8+PEaP368S20FAADAD9zqHksAAABUXQRLAAAAmIJgCQAAAFMQLAEAAGAKt+u8A9yoDqen/ewz6n/OobQDkqS0bw64/KQFPz8/hYY1d23hS+zf/3Wl11EVtgkAuBzBEpCUlXlKf3v7Tf3uDyMU1PCm6779w+lpiu4ZVen1TIkfU6nlN326w+Vw6efnJ0kaPmxYpdpQGWVtAABYg2AJSMrKytSCV/6iu2N+ZUmwLDtTueKvf1XLlq1cWoeXp4eKHSUuLbt//9caPmyYy2dMJSk0rLk2fbrD5XWkfXNAj04ao7kJSWoeHlHh5c064woAcB3BEnAjLVu2UseOHV1atjLB0ixmBLvm4RFqG9nehNYAAK43Ou8AAADAFARLAAAAmIJL4ag2KtOrOu2bA+W+V5Q79Kiu7D2WAABUFsES1YJZvaofneR6r+qq3qP60nYAAOAKgiWqBTN6VbvKHXpUH0o7oCnxYzRvQZKaNa94j2qJXtUAgMojWKJaqUyvaqtVJtTZbKXfm4dHqE1belQDAKxB5x0AAACYgmAJAAAAUxAsAQAAYArusYSkyg3Vcyjth6F6DMO17dNxBACAqo9gCdOG6pkS7/pQPVLlhuu50QUFNdRTTz2loKCGVjcFAHADI1jCeabS37+OPD1d+JWwSZ4eNjlKDMmFM5YOR7FycrIrNVzPjS6o4U2aNWuWDmWcc/msMQAAlUWwhJOnp5fsdnvFF/w+WHq4GCwBAED1QOcdAAAAmIJgCQAAAFMQLAEAAGAKgiUAAABMQbAEAACAKQiWAAAAMAXBEgAAAKYgWAIAAMAUBEsAAACYgmAJAAAAUxAsAQAAYAqCJQAAAExBsAQAAIApCJYAAAAwBcESAAAApiBYAgAAwBQESwBuISiooeInP6agoIZWNwUA4CIvqxsAAJIU1PAmPTxlhtXNAABUAmcsAQAAYAqCJQAAAExBsAQAAIApCJYAAAAwBZ134ORwFLu2oE0q8bDJUWJIxnXc7hXs3/+1aety520CAOCOCJaQn5+fJCknJ9st2lGZZYcPG2ZWc1xuA25MWZmn9NfX5+u+fkPUIOgmq5sDAJYgWEKhYc216dMdys3NdWn5Q2kHNCV+jOYtSFKz5hEurcPPz0+hYc1dWlaq/D5UVmXbj6ovKytTTz/9tDrefhfBEsANi2AJSapUKLLZSr83D49Qm7btTWpRxRHsAACwFp13AAAAYAqCJQAAAExBsAQAAIApCJYAAEmlPdtfnfeisjJPWd0UAFWU2wVLh8OhpKQkxcTEqF27durXr5/WrVtXbp7NmzcrNjZWkZGR6tGjh5599tmf7Q28Z88etWnTRmvWrLnsZ2+88Yaio6MVGRmp/v37a9OmTabvEwBUBVlZmVrwyl+UlZVpdVMAVFFuFyznzZunhIQEDRo0SEuWLNEdd9yh6dOna/369ZKkjRs3Ki4uTj4+Ppo/f76efPJJ7dixQ8OHD1dx8eUDbRcWFmrGjBlX/NnSpUv18ssva8CAAVq4cKFCQkI0adIkbd++/ZrvJwAAQHXjVsMN5eXlKTk5WcOHD9eYMWMkSV27dtXevXuVnJysvn37asGCBQoPD9fSpUvl7e0tSYqKilJ0dLTWrFmjwYMHl1vn/Pnzdf78+cu2VVBQoCVLlmjEiBGaMGGCJKlnz54aMmSIEhMT9eabb17bnQUAAKhm3OqMZY0aNbRy5UqNHDmy3HS73a7CwkJJ0qFDh9S9e3dnqJSk+vXrq1mzZtq8eXO55Xbt2qXk5GT9+c9/vmxbu3fvVk5Oju655x7nNJvNppiYGG3btk0FBQVm7hoAAEC151ZnLL28vNSyZUtJkmEYOn36tNasWaPPPvtMs2fPliTVrVtXGRkZ5ZYrKirSyZMnneFTKj0jOWPGDI0dO1YtWrS4bFtpaWmSpNDQ0HLTQ0JC5HA4dPToUUVEXPkpMjbbD4OC44djwXGxzqU1gDWqw+ugqu8DrwPrUQPrmV2Diq7HrYLlpdavX69p06ZJknr16qX7779fkjRw4EC99tprSkpKUmxsrAoKCjR//nzl5ubKx8fHufycOXPk4+OjsWPH6tSpy3s4ll0e//HznX19fSXpZzsDhTUOcM4HqZZHhJ566ilFtYtQo0YBVjfnhhbWOMDqJtywzmXWliQ1aVBbzZoEWNsYF5XtQ9OgqrsPEq8Dd0ANrGdWDfLy7BWa322DZfv27ZWcnKz09HQlJCRoyJAhevfddxUfHy+Hw6GEhATNnTtXdrtdgwYNUnR0tA4ePChJ+uKLL7Ry5UqtXr1aXl5X3sWSkpIrTjcMQ5Lk4fHTdwmknzgnH5+iSu5h9WGz1dKsWbOUfuKcDmWcs7o5NySbrfRNJP3EOX3/K4zrLOPb887vAVX0dXA867zze1XcB14H1qMG1jO7Bvn5eRWa322DZUhIiEJCQtS5c2cFBwdrxIgR2rBhg/r166epU6cqPj5ex44dU1BQkPz9/TV06FAFBAQoLy9Pjz/+uB566CGFh4eruLjYGSJLSkpUXFwsLy8v+fv7SyrtMFSnTh3ndvPz8yVJtWvX/sm2GYZ4wVwBx8V61MA6Zce9KtegOuyDVPXbXx1QA+uZVYOKrsOtOu+cOXNGa9eu1ZkzZ8pNj4yMlCSdOnVK27Zt09atW1WjRg2Fh4fL399fxcXFSk1NVevWrbVnzx5lZGQoMTFRbdq0UZs2bRQTEyNJeuKJJ9SmTRtJUlhYmCTpyJEj5bZ15MgReXt7Kzg4+FrvLgAAQLXiVmcs8/PzNWPGDE2ePFnjxo1zTt+6daskqUWLFvrwww/1r3/9Sxs3bpTdXnrdPyUlRTk5OYqJiVHLli317rvvllvvt99+q7i4OE2cOFF33nmnJKlDhw7y8fHRhg0b1K5dO0mll8E3btyoLl26lOt1DgAAgF/mVsEyODhYDzzwgBITE+Xh4aHIyEjt2bNHixcvVvfu3dWzZ081atRIq1at0owZMxQbG6vU1FTNmTNHffr0UVRUlKQfznCWOX78uCSpSZMmzp/VqlVLo0aNUmJioux2uzp06KCUlBTt3btXK1asuL47DgAAUA24VbCUpNmzZys0NFQpKSlasGCBGjRooGHDhmn8+PGy2WyKiIjQkiVLNHfuXI0bN06BgYEaN26cxo4dW+FtTZw4UZ6enlq1apWWL1+u8PBwLVq0SJ06dboGewbA3R1OT/vZESF+zqG0A5KktG8OuHxfk5+fn0LDmru2MAC4AZthcHvt1cjLy3MOTfRl6nH5+DDcUBmbTWrWJECHMugFaBVqUHmH09MU3TPK6mZo06c7LAuXe77arQfuv1Pv/XOL2ka2t6QNlcHrwHrUwHpm1yA/P0/tWjSVVDoU4y8Nt+h2ZywBwAplZyr9/evI09OFt0ab5Olhk6PEkFx4M3c4ipWTk+3yGVMAcAcESwC4hKenl7NjYIV8Hyw9XAyWAFAduNVwQwAAAKi6CJYAAAAwBZfCAaAaqUzP9rRvDpT7XlH0agdAsASAasKsnu2PThrj8rJW9moHYD2CJQBUE5Xu2V4J9GoHIBEsAaDacblnOwBUEp13AAAAYAqCJQAAAExBsAQAAIApCJYAAAAwBcESAFAtZGWe0qxZs5SVecrqpgA3LIIlAKBayMrK1NNPP62srEyrmwLcsAiWAAAAMAXBEgAAAKYgWAIAAMAUBEsAAACYgmAJAAAAUxAsAQAAYAqCJQAAAExBsAQAAIApCJYAAAAwBcESAAAApvCyugEA4E4cjmLXFrRJJR42OUoMybiO273G67re2zycnqbc3FyXlj2UdkCSlPbNARku1ECS/Pz8FBrW3LWFARAsAUAqDRSSlJOT7RbtqMyyVu5DZdp/OD1N0T2jKt2GKfFjKrX8pk93EC4BFxEsAUBSaFhzbfp0R6XOlk2JH6N5C5LUrHmES+uo7Nmyyu5D2jcH9OikMZqbkKTm4RXfh8q2v6zd/v515OnpwseTTfKs5FnjnJxsl48fAIIlADhVJhTZbKXfm4dHqE3b9ia1qOLMONPWPDxCbSOt2wdPTy/Z7faKL/h9sPRwMVgCqDw67wAAAMAUBEsAAACYgmAJAAAAUxAsAQAAYAqCJQAAAExBsAQAAIApCJYAAAAwBcESAAAApiBYAgAAwBQESwAwQVBQQz311FMKCmpodVMAwDIESwAwQVDDmzRr1iwFNbzJ6qYAgGUIlgAAADAFwRIAAACmIFgCAADAFARLAAAAmIJgCQAAAFMQLAEAkkqHTIqf/BhDJlkoK/OUXp33orIyT1ndFMAlBEsAgKTSIZMenjKDIZMslJWVqQWv/EVZWZlWNwVwCcESAAAApiBYAgAAwBQESwAAAJiCYAkAAABTECwBAABgCrcLlg6HQ0lJSYqJiVG7du3Ur18/rVu3rtw8mzdvVmxsrCIjI9WjRw89++yzys3NLTfP/v37NXr0aEVFRem2227TY489pqysrMu298Ybbyg6OlqRkZHq37+/Nm3adE33DwAAoLpyu2A5b948JSQkaNCgQVqyZInuuOMOTZ8+XevXr5ckbdy4UXFxcfLx8dH8+fP15JNPaseOHRo+fLiKi4slSadOndLw4cOVm5urOXPmaNasWdq1a5dGjhzpnEeSli5dqpdfflkDBgzQwoULFRISokmTJmn79u2W7DsAAEBV5mV1Ay6Vl5en5ORkDR8+XGPGjJEkde3aVXv37lVycrL69u2rBQsWKDw8XEuXLpW3t7ckKSoqStHR0VqzZo0GDx6sv//977pw4YJee+01BQQESJLq1aunYcOG6fPPP1ePHj1UUFCgJUuWaMSIEZowYYIkqWfPnhoyZIgSExP15ptvWnEIAAAAqiy3OmNZo0YNrVy5UiNHjiw33W63q7CwUJJ06NAhde/e3RkqJal+/fpq1qyZNm/eLEkaNmyY3n77bWeoLFuHJOd6du/erZycHN1zzz3OeWw2m2JiYrRt2zYVFBRck30EAACortzqjKWXl5datmwpSTIMQ6dPn9aaNWv02Wefafbs2ZKkunXrKiMjo9xyRUVFOnnypDM01qtXT/Xq1ZMkXbx4Ufv27dMzzzyj0NBQde/eXZKUlpYmSQoNDS23rpCQEDkcDh09elQRERHXbF8BAACqG7cKlpdav369pk2bJknq1auX7r//fknSwIED9dprrykpKUmxsbEqKCjQ/PnzlZubKx8fn8vW07dvXx05ckQ1atRQQkKCatSoIUk6f/68JMnPz6/c/L6+vpJ0WWegS9lspV8oVXYsOCbWoQbWowaVV3bsHI5iycXjWOJhk6PEcGlZh6PY2Y7K1DH9UJry8n76M+TnHEo74PzuSht8ff0U1qy5S9s2A68D65ldg4qux22DZfv27ZWcnKz09HQlJCRoyJAhevfddxUfHy+Hw6GEhATNnTtXdrtdgwYNUnR0tA4ePHjZep566ilJ0po1axQXF6cXX3xR/fv3V0lJyRW3axilb0geHj99l0BY4wBnAMUPwhoHWN2EGx41sB41cJ0jv7EkKScn29J2tGreWM2aBLi07MGDBxXdM6rSbZgSP8blZQ8cOKBbbrml0m2oDF4H1jOrBnl59grN77bBMiQkRCEhIercubOCg4M1YsQIbdiwQf369dPUqVMVHx+vY8eOKSgoSP7+/ho6dGi5eyrLdOvWzfk9MzNTiYmJ6t+/v/z9/SWVdhiqU6eOc/78/HxJUu3atX+ybeknzsnHp8jEva3abLbSX+D0E+dkuHaiAJVEDaxHDSrP06eBNn26w+WzfTab1KRBbWV8e97lGvj6+snTp4EOZZxzafmv005Ikvz968jT6/p+xDqKi5WTk62v007I06fBdd12GV4H1jO7Bvn5eRWa362C5ZkzZ/Tpp5+qZ8+eql+/vnN6ZGSkpNJhhLZt26aLFy+qR48eCg8PlyQVFxcrNTVVAwcOlCR9/vnnKiwsVK9evcqtv23btnr77bclSWFhYZKkI0eOqF27ds55jhw5Im9vbwUHB/9kOw1DvGCugONiPWpgPWpQOaFhrl/GtdmkZk0CFJBRuQ9UM5b19PSS3atiZ3oqzfihDVb/DrpDG250ZtWgoutwq17h+fn5mjFjhlavXl1u+tatWyVJLVq00IcffqiZM2eqqOiHM4YpKSnKyclRTEyMJGnt2rWaPn16ufski4uL9fnnnzs7B3Xo0EE+Pj7asGGDcx7DMLRx40Z16dKlXK9zAAAA/DK3OmMZHBysBx54QImJifLw8FBkZKT27NmjxYsXq3v37urZs6caNWqkVatWacaMGYqNjVVqaqrmzJmjPn36KCqq9L6W0aNH66OPPtKYMWM0evRoGYaht956S2lpaVq+fLkkqVatWho1apQSExNlt9vVoUMHpaSkaO/evVqxYoWVhwEAAKBKcqtgKUmzZ89WaGioUlJStGDBAjVo0EDDhg3T+PHjZbPZFBERoSVLlmju3LkaN26cAgMDNW7cOI0dO9a5joiICL399tuaN2+eHn/8cRUWFqpDhw5KTk7Wrbfe6pxv4sSJ8vT01KpVq7R8+XKFh4dr0aJF6tSpkwV7DgAAULW5XbD09vZWXFyc4uLifnKebt26OTvl/JQ2bdpo2bJlPzuPzWbT+PHjNX78eJfaCgAAgB+41T2WAAAAqLoIlgAAADAFwRIAAACmIFgCAADAFARLAAAAmIJgCQAAAFMQLAEAAGAK04Ll6dOntW/fPmVnZ5u1SgAAAFQhlRogvaioSMuWLdOqVat08uRJSdLChQt18eJFvf/++3riiScUHBxsSkMBAADg3lwOlhcvXtTIkSO1a9cuGYYhqfRJNpJ04MABbdmyRfv27dPKlSvVqFEjc1oLAAAAt+XypfDXX39dO3fulCTdfffd5X6Wn58vSfr222+1ePHiSjQPAAAAVYXLwfKDDz6QzWbTkCFDlJiYWO5nf/rTnzRo0CAZhqH//Oc/lW4kAAAA3J/LwTIjI0OS1Llz5yv+/Pbbb5dUetYSAAAA1Z/LwdLX11eSdOjQoSv+/L///a8kKSAgwNVNAAAAoApxufPOHXfcoQ8++EDLli2Tt7e3c/q2bdu0efNmpaSkyGaz6bbbbjOloQAAAHBvLgfLRx55RP/+97+Vk5OjV155xdkj/K9//askyTAM1apVS+PHjzenpQAAAHBrLl8KDw4O1ptvvqlbbrlFhmFc9hUSEqI33nhDYWFhZrYXAAAAbqpSA6S3atVK77//vr788kt99dVXysnJka+vr1q3bq1OnTo5z2ICAACg+qtUsCwTGhqqdu3aOf+/Z88eQiUAAMANplLPCv/iiy/0q1/9SosWLSo3/Y9//KN69+6t7du3V6pxAAAAqDpcDpZfffWVRo8ercOHD+vrr792Tj9+/LguXLigEydOaNSoUdq/f78pDQUAAIB7czlYvvbaayoqKpKHh4ciIyOd0318fDRixAh5eXmpuLiYRzoCAADcIFwOlmX3UY4bN05Tp051Tq9Xr55mzJihsWPHyjAMffXVV6Y0FAAAAO7N5WB59uxZSVJISMgVfx4cHCxJOn36tKubAAAAQBXicrAMCgqSJH300UdX/Pk//vEPSVJgYKCrmwAAAEAV4vJwQ3feeafefvttffzxxxo4cKB69OihunXr6rvvvtPWrVv19ddfy2az6a677jKzvQAAAHBTLgfLuLg4bdy4UVlZWfr666/L9QyXSh/p2LBhQ8XFxVW6kQAAAHB/Ll8KDwwM1N/+9jd169btio907N69u9555x0uhQMAANwgKvXknSZNmmjZsmXKzMzUvn37dP78edWuXVutW7dWw4YNzWojAAAAqgBTHunYsGFDgiQAAMAN7qqD5V//+ldJ0l133aXg4GDn/6/GsGHDKt4yAAAAVClXHSyff/552Ww2NWnSRMHBwc7/Xw2CJQAAQPVXqUvhhmH84jxXGz4BAABQtV11sHzhhRckSW3atCn3fwAAAECqQLAcMGBAuf+X9f5u3Lix6Y0CAABA1ePyOJZPPvmk7r77br3yyitmtgcAAABVlMvBsqCgQJIUHh5uWmMAAABQdbkcLO+9914ZhqFPPvlEJSUlZrYJAAAAVZDLvcI7d+6sL774Qh988IG++OIL3Xrrrapbt65q1KghD4/yefXxxx+vdEMBAADg3lwOlk8++aRzKKHTp09r06ZNPzkvwRIAAKD6YxxLAAAAmMLlYFmRRzoCAACg+qtwsDx79qw+//xznTx5Un5+furSpYuaNWt2LdoGAACAKqRCwXLFihWaP3++c6ihMgMGDNAzzzwjL69KXVkHAABAFXbVww199NFHeuGFF3ThwgUZhlHua+3atfrLX/5yLdsJAAAAN3fVwXLFihWSSjvjtG3bVqNHj9Z9990nDw8PGYahlStXKj8//5o1FAAAAO7tqq9dHzp0SDabTV27dtWyZcucvb3feustPffccyoqKlJ6erratGlzzRoLAAAA93XVZyxzc3MlSX369Ck3hFCfPn2c/z537px5LQMAAECVctXBsri4WJLk7+9fbnrdunWd/y4sLDSpWQAAAKhqrjpYlg2G/uPHNV569pJnhgMAANy4Kjw+0H//+1+dP3++Qj974IEHKtwwAAAAVC0VDpbJycmXTSs7a/lTP6tIsHQ4HFq2bJlWr16tzMxMhYaG6sEHH1T//v2d82zevFmJiYlKTU1VQECA7r33Xj3yyCPy8/NzznPs2DHNnTtXO3bs0IULF3TLLbfo4YcfVteuXctt74033tDbb7+tzMxMNWvWTPHx8YqOjr7q9gIAAKDUVV8Kl3TZ+JVX+1UR8+bNU0JCggYNGqQlS5bojjvu0PTp07V+/XpJ0saNGxUXFycfHx/Nnz9fTz75pHbs2KHhw4c77wM9d+6chg4dqrS0NP3pT3/SK6+8oqCgII0aNUrbtm1zbmvp0qV6+eWXNWDAAC1cuFAhISGaNGmStm/fXqE2AwAAoAJnLCdOnHgt2yFJysvLU3JysoYPH64xY8ZIkrp27aq9e/cqOTlZffv21YIFCxQeHq6lS5fK29tbkhQVFaXo6GitWbNGgwcP1tq1a3XmzBmtWrVKDRs2lCR1795d/fv317Jly9SlSxcVFBRoyZIlGjFihCZMmCBJ6tmzp4YMGaLExES9+eab13x/AQAAqhO3CpY1atTQypUrFRgYWG663W53Dnd06NAhDR061BkqJal+/fpq1qyZNm/erMGDB6thw4YaMWKEM1RKpZ2Obr75Zh06dEiStHv3buXk5Oiee+5xzmOz2RQTE6N58+apoKBANWvWvJa7CwAAUK241cO9vby81LJlS0mll91Pnz6tNWvW6LPPPtPs2bMllQ5vlJGRUW65oqIinTx50jnc0f3336/777+/3Dznzp3Ttm3bnPdYpqWlSZJCQ0PLzRcSEiKHw6GjR48qIiLC9H0EAACortwqWF5q/fr1mjZtmiSpV69ezqA4cOBAvfbaa0pKSlJsbKwKCgo0f/585ebmysfH54rrcjgceuKJJ5Sfn6+HHnpIkpy91y/t8CNJvr6+kn4YEP5KbLbSL5QqOxYcE+tQA+tRA+u5Qw2c27Z9/3VdN/5DG6w6Bu5Qgxud2TWo6HrcNli2b99eycnJSk9PV0JCgoYMGaJ3331X8fHxcjgcSkhI0Ny5c2W32zVo0CBFR0fr4MGDl62nqKhIjz32mDZt2qRZs2YpMjJS0k+PuflT43VeKqxxgDOA4gdhjQOsbsINjxpYjxpYz8oanMusLUny9LDJ0+P6pquS77fXNKi2mjUJuK7b/jFeB9YzqwZ5efYKze+2wTIkJEQhISHq3LmzgoODNWLECG3YsEH9+vXT1KlTFR8fr2PHjikoKEj+/v4aOnSoAgICyq0jOztbEydO1Pbt2/XnP/9Zv/vd75w/K3uCUF5enurUqeOcnp+fL0mqXbv2T7Yt/cQ5+fgUmbi3VZvNVvoLnH7inCo4CABMQg2sRw2s5w41OJ5VejXMUWLIo+T6NsLx/faOZ51XQMa567rtMu5Qgxud2TXIz8+r0PxuFSzPnDmjTz/9VD179lT9+vWd08vOMp46dUrbtm3TxYsX1aNHD4WHh0sqfdxkamqqBg4c6Fzm5MmTGjVqlI4fP6558+Zdds9lWFiYJOnIkSNq166dc/qRI0fk7e2t4ODgn2ynYYgXzBVwXKxHDaxHDaxnZQ2c2zW+/7quG/+hDVb/DrpDG250ZtWgouuo0DiW11p+fr5mzJih1atXl5u+detWSVKLFi304YcfaubMmSoq+uGMYUpKinJychQTEyOp9P7IESNGKCsrS8uXL78sVEpShw4d5OPjow0bNjinGYahjRs3qkuXLuV6nQMAAOCXudUZy+DgYD3wwANKTEyUh4eHIiMjtWfPHi1evFjdu3dXz5491ahRI61atUozZsxQbGysUlNTNWfOHPXp00dRUVGSpISEBB0+fFjx8fGy2+363//+59yGt7e3WrdurVq1amnUqFFKTEyU3W5Xhw4dlJKSor1792rFihUWHQEAAICqy62CpSTNnj1boaGhSklJ0YIFC9SgQQMNGzZM48ePl81mU0REhJYsWaK5c+dq3LhxCgwM1Lhx4zR27FjnOj766CNJ0oIFC7RgwYJy62/SpIn+9a9/SSodm9PT01OrVq3S8uXLFR4erkWLFqlTp07Xb4cBAACqCbcLlt7e3oqLi1NcXNxPztOtWzd169btJ3++ZcuWq9qWzWbT+PHjNX78+Io2EwAA/EhW5in99fX5uq/fEDUIusnq5sACbnWPJQAAqLqysjL19NNPKysr0+qmwCIESwAAAJiCYAkAAABTECwBAABgCoIlAAAATEGwBAAAgCkIlgAAADAFwRIAAACmIFgCAADAFARLAAAAmIJgCQAAAFMQLAEAAGAKgiUAAABMQbAEAACAKQiWAAAAMAXBEgAAAKYgWAIAAMAUBEsAAACYgmAJAAAAUxAsAQAAYAqCJQAAAExBsAQAAIApCJYAAAAwBcESAAAApiBYAgAAwBQESwAAAJiCYAkAAABTECwBAABgCoIlAAAATEGwBAAAgCkIlgAAADAFwRIAAACmIFgCAADAFARLAAAAmIJgCQAAAFMQLAEAAGAKgiUAAABMQbAEAACAKQiWAAAAMAXBEgAAAKYgWAIAAMAUBEsAAACYgmAJAAAAU3hZ3QAAAKobh6O4ym7zcHqacnNzXVr2UNoBSVLaNwdkGK5t38/PT6FhzV1bGJYjWAIAYBI/Pz9JUk5OtuVtcMXh9DRF94yqdBumxI+p1PKbPt1BuKyiCJYAAJgkNKy5Nn26w+UzfpVV2bN9Ze32968jT08XIoJN8vSwyVFiSC6csXQ4ipWTk23Z8UPlESwBADBRdTjT5unpJbvdXvEFvw+WHi4GS1R9dN4BAACAKQiWAAAAMAXBEgAAAKYgWAIAAMAUbhcsHQ6HkpKSFBMTo3bt2qlfv35at25duXk2b96s2NhYRUZGqkePHnr22Wd/tgfZCy+8oD/+8Y9X/Nkbb7yh6OhoRUZGqn///tq0aZOp+wMAAHCjcLtgOW/ePCUkJGjQoEFasmSJ7rjjDk2fPl3r16+XJG3cuFFxcXHy8fHR/Pnz9eSTT2rHjh0aPny4iosvHxw2KSlJb7755hW3tXTpUr388ssaMGCAFi5cqJCQEE2aNEnbt2+/lrsIAABQLbnVcEN5eXlKTk7W8OHDNWZM6eCqXbt21d69e5WcnKy+fftqwYIFCg8P19KlS+Xt7S1JioqKUnR0tNasWaPBgwdLko4dO6YXX3xRmzdvVu3atS/bVkFBgZYsWaIRI0ZowoQJkqSePXtqyJAhSkxM/MkwCgAAgCtzqzOWNWrU0MqVKzVy5Mhy0+12uwoLCyVJhw4dUvfu3Z2hUpLq16+vZs2aafPmzc5pL7zwgo4ePaoVK1aoVatWl21r9+7dysnJ0T333OOcZrPZFBMTo23btqmgoMDs3QMAAKjW3OqMpZeXl1q2bClJMgxDp0+f1po1a/TZZ59p9uzZkqS6desqIyOj3HJFRUU6efKkM3xK0iOPPKJbbrlFNpvtittKS0uTJIWGhpabHhISIofDoaNHjyoiIsKsXQMAAKj23CpYXmr9+vWaNm2aJKlXr166//77JUkDBw7Ua6+9pqSkJMXGxqqgoEDz589Xbm6ufHx8nMv/Uig8f/68pMufqerr6ytJP9sZyGYr/UKpsmPBMbEONbAeNbAeNag857Gzff9VqZW5vgyfs64z+3VQ0fW4bbBs3769kpOTlZ6eroSEBA0ZMkTvvvuu4uPj5XA4lJCQoLlz58put2vQoEGKjo7WwYMHr3r9JSUlV5xuGKXPoPLw+Om7BMIaBzgDKH4Q1jjA6ibc8KiB9aiB9aiB685llvZJ8PSwydPD9WTi6rIl3y/XNKi2mjUJcHn7MO91kJdXsUd7um2wDAkJUUhIiDp37qzg4GCNGDFCGzZsUL9+/TR16lTFx8fr2LFjCgoKkr+/v4YOHaqAgICrXr+/v7+k0g5DderUcU7Pz8+XpCt2+CmTfuKcfHyKXNuxashmK/0FTj9xTgbPhrUENbAeNbAeNai841mlV/McJUbp875d4Olhk8PFZcuWO551XgEZ51xax43O7NdBfn5eheZ3q2B55swZffrpp+rZs6fq16/vnB4ZGSlJOnXqlLZt26aLFy+qR48eCg8PlyQVFxcrNTVVAwcOvOpthYWFSZKOHDmidu3aOacfOXJE3t7eCg4O/sllDUO8aV0Bx8V61MB61MB61MB1zuNmfP9VUZeeqHRleeOHdlDDyjHrGFZ0HW7VKzw/P18zZszQ6tWry03funWrJKlFixb68MMPNXPmTBUV/XDGMCUlRTk5OYqJibnqbXXo0EE+Pj7asGGDc5phGNq4caO6dOlSrtc5AAAAfplbnbEMDg7WAw88oMTERHl4eCgyMlJ79uzR4sWL1b17d/Xs2VONGjXSqlWrNGPGDMXGxio1NVVz5sxRnz59FBUVddXbqlWrlkaNGqXExETZ7XZ16NBBKSkp2rt3r1asWHEN9xIAAKB6cqtgKUmzZ89WaGioUlJStGDBAjVo0EDDhg3T+PHjZbPZFBERoSVLlmju3LkaN26cAgMDNW7cOI0dO7bC25o4caI8PT21atUqLV++XOHh4Vq0aJE6dep0DfYMAACgenO7YOnt7a24uDjFxcX95DzdunVTt27drnqdb7311hWn22w2jR8/XuPHj69wOwEAAFCeW91jCQAAgKqLYAkAAABTECwBAABgCoIlAAAATEGwBAAAgCkIlgAAADAFwRIAAOB7WZmn9Oq8F5WVecrqplRJBEsAAIDvZWVlasErf1FWVqbVTamSCJYAAAAwBcESAAAApiBYAgAAwBQESwAAAJiCYAkAAABTECwBAABgCoIlAAAATEGwBAAAgCkIlgAAADAFwRIAAACmIFgCAADAFARLAAAAmIJgCQAAAFMQLAEAAGAKgiUAAABMQbAEAACAKQiWAAAAMAXBEgAAAKYgWAIAAMAUBEsAAACYgmAJAAAAUxAsAQAAYAqCJQAAAExBsAQAAIApCJYAAAAwBcESAAAApiBYAgAAwBQESwAAAJiCYAkAAABTECwBAABgCoIlAAAATEGwBAAAgCkIlgAAADAFwRIAAACmIFgCAADAFF5WNwAAALgXh6PYtQVtUomHTY4SQzKu43Z/5HB6mnJzc11aNu2bA+W+V5Sfn59Cw5q7tGx1QLAEAACSSkORJOXkZLtFO1xxOD1N0T2jKt2GRyeNcXnZTZ/uuGHDJcESAABIkkLDmmvTpztcPtt3KO2ApsSP0bwFSWrWPMKldVT2jF9Z2/3968jT8/rGHIejWDk52S4fv+qAYAkAAJwqE+psttLvzcMj1KZte5Na5BpPTy/Z7XZL23AjovMOAAAATEGwBAAAgCkIlgAAADAF91gCAIBqx6yhi9x9m+6GYAkAAKoNdxgyqTLDJVV1bhcsHQ6Hli1bptWrVyszM1OhoaF68MEH1b9/f+c8mzdvVmJiolJTUxUQEKB7771XjzzySLlC5ubm6qWXXtLHH3+svLw8dejQQU888YTCw8PLbe+NN97Q22+/rczMTDVr1kzx8fGKjo6+bvsLAADMU9khk9K+OaBHJ43R3IQkNQ+v+JBJDJDuZubNm6cVK1Zo0qRJioyM1CeffKLp06fLw8NDffv21caNGxUfH68uXbpo/vz5Ki4u1uLFizV8+HCtXLlSXl6lu/Too4/qyy+/1LRp0+Tn56eFCxdq+PDh+uCDDxQQECBJWrp0qebNm6cJEyaobdu2SklJ0aRJk7RixQp17tzZwqMAAABcZUawax4eobaR1g6ZVBW5VbDMy8tTcnKyhg8frjFjSke879q1q/bu3avk5GT17dtXCxYsUHh4uJYuXSpvb29JUlRUlKKjo7VmzRoNHjxYu3bt0pYtW5SUlKRevXo557n77rv1zjvvaPz48SooKNCSJUs0YsQITZgwQZLUs2dPDRkyRImJiXrzzTctOQYAAABVlVv1Cq9Ro4ZWrlypkSNHlptut9tVWFgoSTp06JC6d+/uDJWSVL9+fTVr1kybN2+WJP373/+Wj4+PunXr5pynXr166ty5sz799FNJ0u7du5WTk6N77rnHOY/NZlNMTIy2bdumgoKCa7afAAAA1ZFbnbH08vJSy5YtJUmGYej06dNas2aNPvvsM82ePVuSVLduXWVkZJRbrqioSCdPnnSGz7S0NDVt2tR5WbzMzTffrPXr1zvnkaTQ0NBy84SEhMjhcOjo0aOKiLjyvRU22w9PF8APx4JjYh1qYD1qYD1qYL1La1BV61DV98Hs10FF1+NWwfJS69ev17Rp0yRJvXr10v333y9JGjhwoF577TUlJSUpNjZWBQUFmj9/vnJzc+Xj4yNJOn/+/BV7ZPn6+iovL885j3R5zy1fX19J+tmbfsMaBzjnww/CGgdY3YQbHjWwHjWwHjWwTi2PCD311FOKahehRo0CrG6OS85l1pYkNQ2qrWZNAqxtTCWY9TrIy6vYYzHdNli2b99eycnJSk9PV0JCgoYMGaJ3331X8fHxcjgcSkhI0Ny5c2W32zVo0CBFR0fr4MGDkqSSkhLZfiJie3h4OOe5EsMwys13JeknzsnHp6gyu1et2Gylv8DpJ87p+8OH64waWI8aWI8aWM9mq6VZs2Yp/cQ5Hco4Z3VzXHI867zze0AV3AezXwf5+XkVmt9tg2VISIhCQkLUuXNnBQcHa8SIEdqwYYP69eunqVOnKj4+XseOHVNQUJD8/f01dOhQZ2/v2rVr68yZM5etMy8vT7Vrl/4l4u/v75xWp04d5zz5+fnOdfwUwxBvWlfAcbEeNbAeNbAeNbBeVa5BWbur8j5I5rW/outwq847Z86c0dq1ay8LhZGRkZKkU6dOadu2bdq6datq1Kih8PBw+fv7q7i4WKmpqWrdurUkKSwsTMePH7/srOTRo0fVrFkz5zySdOTIkXLzHDlyRN7e3goODr4m+wgAAFBduVWwzM/P14wZM7R69epy07du3SpJatGihT788EPNnDlTRUU/XIpOSUlRTk6OYmJiJEndu3dXXl6eczlJOnv2rLZv367u3btLkjp06CAfHx9t2LDBOY9hGNq4caO6dOlSrtc5AAAAfplbXQoPDg7WAw88oMTERHl4eCgyMlJ79uzR4sWL1b17d/Xs2VONGjXSqlWrNGPGDMXGxio1NVVz5sxRnz59FBUVJUnq3LmzunTpomnTpmnatGkKCAjQggULVLt2bQ0ZMkSSVKtWLY0aNUqJiYmy2+3q0KGDUlJStHfvXq1YscLKwwAAAFAluVWwlKTZs2crNDRUKSkpWrBggRo0aKBhw4Zp/PjxstlsioiI0JIlSzR37lyNGzdOgYGBGjdunMaOHVtuPQsXLtSLL76ol156SSUlJerYsaPmz59f7n7KiRMnytPTU6tWrdLy5csVHh6uRYsWqVOnTtd7twEAAKo8m2FU5VtTr5+8vDzn0ERfph6Xjw/DDZWx2aRmTQJ0KIOemFahBtajBtajBtarDjXY89VuPXD/nXrvn1uq5CMdza5Bfn6e2rVoKql0KMZfGm7Rre6xBAAAQNVFsAQAAIApCJYAAAAwBcESAAAApiBYAgAAwBQESwAAAJiCYAkAAABTECwBAABgCoIlAAAATEGwBAAAgCkIlgAAADAFwRIAAACmIFgCAADAFARLAAAAmIJgCQAAAFMQLAEAAGAKgiUAAABMQbAEAACAKQiWAAAA3wsKaqj4yY8pKKih1U2pkrysbgAAAIC7CGp4kx6eMsPqZlRZnLEEAACAKQiWAAAAMAXBEgAAAKYgWAIAAMAUBEsAAACYgmAJAAAAUxAsAQAAYAqCJQAAAExBsAQAAIApCJYAAAAwBcESAAAApiBYAgAAwBQESwAAAJiCYAkAAABTeFndgKrCMAznv/Pz8y1sifux2aS8PLvy8/N0yWHCdUQNrEcNrEcNrEcNrGd2DS7NPMZVrJBgeZUuPbC3d4iwsCUAAADXX35+vvz8/H52Hi6FAwAAwBQ242rOa0IlJSU6ffq0JMnHx0c2m83iFgEAAFxbhmE4r9oGBgbKw+Pnz0kSLAEAAGAKLoUDAADAFARLAAAAmIJgCQAAAFMQLAEAAGAKxrFElXT48GHt379fp06dUrdu3RQYGKi6deuqpKTkF3uswRzUwHrUwHrUwHrUwL3QKxxVzs6dOzVhwgR5enoqJydHNWrUUKdOnTRp0iS1bt2aN5PrgBpYjxpYjxpYjxq4H442qpTMzEw98cQT+vWvf63ly5fryy+/1JgxY3Tq1Ck9+OCD+r//+z95eHiopKTE6qZWW9TAetTAetTAetTAPREsUaWcOXNG2dnZuuuuuxQRUfpozYceekiTJ09WSEiIRo8erZ07d8rDw+OqnmmKiqMG1qMG1qMG1qMG7olgiSolOztbFy5cUN26dSVJFy9elCT16tVLjzzyiFq0aKEpU6YoNTWVpyNdI+fOnaMGFuN1YD1qYD1q4J4IlqhSOnbsqMDAQC1YsECSVKNGDRUVFUmSbr/9dj300EOqU6eO3nzzTeXl5VnZ1Grlu+++cx7nTp06UQOLdezYUfXr16cG11HZGa+y77wXWY8auCeCJdzakSNHdOHCBUmlz2uvUaOGRowYof/7v//Tq6++Kkmy2+3ON5O7775bvXv31pYtW1RQUGBZu6uT//73v+rbt6/S0tJkGAY1sEBubq5ycnJ09uxZORwO1ahRQ6NGjdL27dupwXXicDgkScXFxZIkLy8vXgfXGZ8HVQPBEm7r/PnzGjx4sCZPnqzc3Fxnz77o6Gh1795d7733npYsWSKp9M2k7I3jt7/9rb777jvt2bPHsrZXFzt27NDYsWPVu3dvhYaGOi8nUYPr58svv1RcXJz69++vu+++W7NmzdKBAwfUr18/9ezZkxpcB3v37tXTTz+tkSNH6uGHH9b+/fvl6emp3r178zq4Tvg8qDoIlnBrDRo00JYtWzRz5kznpYyGDRsqPj5eERER+vvf/6558+ZJkmrWrClJSk1NVWBgoOrXr29Zu6uDHTt2aMSIERoyZIimT5/uPL5SaQ0mTpyoiIgIvfPOO9TgGvnqq6/00EMP6eabb9ZDDz2ksWPH6r333tPChQvl4+OjCRMm8Dq4xnbv3q1hw4YpOztbvr6+On78uH7729/q0KFDatSokfO96G9/+xs1uMb4PKgaGMcSbu25557Tp59+quzsbLVt21avvvqqfH19JUnp6elatGiRPv74Y3Xs2FGDBw/Wd999p7Vr18put2vFihWMX+aiAwcOKDY2Vn379tVTTz0lb29vSdI333yj3Nxc1axZUy1atNDZs2f10ksvaePGjdTAZIWFhZo6daok6dlnn5W/v78k6fXXX9e8efOUkpKi1q1b68iRI1q4cKG2bNmi9u3bUwMT5efnKz4+XnXr1tVTTz2l2rVra/fu3Ro/frwmTpyo3/3ud5Kk48ePKyEhQZs2beJ1cA3xeVA1cJTh1vz9/VWzZk098cQT2rlzpx599FHl5ORIksLCwjR79mzNnDlT3377raZOnapFixbJz89Py5cvl4eHh/O+KFTMoUOHVFhYqJtvvtkZKl955RWNGDFCv//97/XAAw9o2LBhOnLkiLMGWVlZ1MBEhYWF+uabbxQRESF/f385HA4ZhqEePXrIMAx9/fXXkqSQkBA99thjevzxx3kdmKygoEBHjx5V27ZtVbt2bUlS06ZN5e3trUOHDumVV17RX//6V128eFHPPvus/vznPyszM5MaXCN8HlQRBuCGSkpKDMMwjM8++8z4/e9/b2RnZxvvvvuu0aFDB2PixInG2bNnjTFjxhi7d+92LpOammqcOHHCcDgchmEYRlFRkSVtry5efvll49ZbbzWOHj1qrF271ujQoYPxxhtvGJ999pnxt7/9zbjrrruMwYMHG998841zGWpgnuLiYmPgwIHGww8/XG56fn6+0aZNG+ONN94wDMNwHusyX3/9NTUwSX5+vnHXXXcZTzzxhHPali1bjDZt2hj333+/cddddxlt27Y1+vbta2zfvt05D68Dc/F5ULXwrHC4pbJOIk2bNtW+fft07Ngx3XvvvfLx8dGTTz6p3r17Kzw8XE2bNpXD4ZCnp6dzgFyptMeglxe/3q4oewTahAkTtHv3bg0fPlwRERGaOXOmBgwYIEnq2rWr6tatqxkzZuhf//qXmjdvLknUwEQeHh7q0qWLvv32W2VkZKhJkyZyOBzOXrG1atVyzidJFy5cUK1atdSyZUvnOqhB5dSoUUO9evXSbbfdpuLiYnl5eemf//ynevfurUcffVQhISHatGmTZs2apffff19RUVGSeB2Yjc+DqoVL4XAbP75MUVRUpAYNGigkJEQnT56Un5+fbrvtNvn6+qqoqEh+fn7y8fGRp6fnZeviXhrXOBwO57GrVauWxowZIy8vL23dulWhoaGSSsfxMwxD9957r+688059+OGHzmmXogauKXsd2Gw2TZ48WaNHj1aTJk0kSZ6eniooKFBxcXG5zlQnT57UK6+8clnPV2rgmrIaeHh46Mknn9R9993nDCYzZ87UM888o5CQEEmlvZJjYmL073//+4pD2lAD1/B5UHVxtOEWDMOQp6en8vPztW/fPkmlQ0bUrFlTjRs31r59+3Tx4kUNGzZMDRs21KOPPqq9e/dqxIgRjE9mkktrsHfvXkmlZyZ79+6toUOH6pZbbpFUGnhsNpuKiop08eJFBQUFOaehci6twVdffSVvb2/nGciy4F72gWu32yVJGRkZmjhxov773/+qVatW1jS8Gvnx6+DSoGIYhvz8/BQQECCp9ExY2VnkkJCQcmEfruPzoGrj3DDcgs1m08WLFzVw4EB17dpVTzzxhDw8POTh4aHmzZtr586dio2Nla+vrxYtWiQfHx95e3vr448/dnYuQeX8uAbh4eGqUaOGpk2bpqKiItWsWVNnz55VvXr1JEknTpxQRkaGevToYXHLq48f16BVq1bOM2Vlwf3s2bOSpDp16igzM1NTpkzRxYsXtXbtWnl6ejpvZYBrflyDFi1aXFaD7Oxs1alTRx4eHjp69KgOHDigTp06WdnsaoXPg6qN4YZgiW+//VbZ2dnKzs5WaGioc4yx9evX6/bbb1eDBg2c827fvl1//OMfdfvtt+uFF15Qo0aNJJXeU1azZk3ZbDY+TF1QkRpI0v79+7Vo0SIVFhaqYcOG+t///icPDw+tXr2a+5dclJ2dLcMwdO7cOTVu3Nj5ofj++++ra9eul9VAKh1fdOjQoXruuef03nvv6cyZM1q3bp3sdrvzPkBcvYrWIC0tTUlJSbpw4YIaNGignTt3SpLzdWAYBmfvK6giNeDzwP0RLHHd/e9//9OTTz6p7Oxsffvtt2rdurV+85vf6A9/+MMV58/MzNSnn36q7t27O99ELsUbecVVtAZS6Rv39OnTdfjwYQUEBOiWW27Rn/70J3l5eTlvmMfV+9///qeXX35Z586d09GjR9WxY0fdcccdGjt2rCT95DE9duyY+vTpo8LCQoWHhzvH6SNUVpwrNSgoKNCjjz6qo0ePql69egoPD9fjjz/O68BFFa0Bnwfuj2CJ62r//v0aNWqU+vTpozvvvFPe3t568cUXlZubqzlz5igyMvKKy/EXqHlcqUHZm3VJSYkKCgrk6empGjVqSBKBxgVff/21fv/73ys2NlY9e/aUr6+vli9fri1btui3v/2tZs6cKenKx9bhcCgmJka+vr5au3atvLy8qIELXKnBpa+DixcvytPT03l2jRpUnKuvAz4P3BuvAlxX7733npo3b664uDjnvXqPP/64hg4dqv379/9ksCwuLubeGZO4UoOye55q1KghHx8f53TDMPgwrSDDMLRq1SqFhYVp7NixCgwMlCQ1atRIBw8e1Ntvv62CggI999xzl50FKxtSaNmyZbr55pvl6elJoHGBqzW49HVQNtxT2fqoQcVU5nXA54F7I/LjuikpKVFqaqr8/f1Vr149GYahoqIitWnTRoGBgUpNTb3icidPntSsWbO0f//+69zi6qcyNXj66acvqwGXnCrO4XDo66+/VkBAgAIDA2UYhgoLC9WoUSO1b99ePXr00L/+9S+99NJLkuT8MD158qRmz56t/fv3KywsTJ6ennI4HAQaF1SmBrwOzFGZGvB54N4IlrimLr3TwsPDQ8HBwTpw4ICysrJks9lkt9tVq1Yt+fj46Pz581dcxxdffKF9+/YpPDz8ejW7WqEG1ru0Bl5eXmrcuLGOHTumEydOyGazOc++nDx5UnfeeadzfNCtW7c6l9u+fbv27t1brgbcz3f1zKoBrwPXUYMbA8ES19SP/5Lv1auXoqKinEOmlI3JZxiG8569Mnl5eZKkBx54QGvWrHFeDkHFUAPr/bgGHTp00NmzZzV37lxt375dO3fu1MiRI5WVlaU//OEPGjdunPLz8/Xvf//buUy/fv303nvvUQMXmVWDsvtaqUHFUYMbA9dQcM0cOHBA7733ntLT09WkSRPFxsbq7rvvVvv27Z330xiGoYKCAuXm5srPz8+57IkTJ/TMM8/o3nvv1YABA+Th4eEcNBdXjxpYLy0tTVu2bFFubq4aN26sQYMG6Y9//KNOnz6tlJQUffDBB/Lx8VHTpk21ZMkSSVJISIjuuecebdmyRZMnT5aHh4e8vb1ls9mogQuogfWowY2DYIlrYteuXYqPj1doaKi8vb31z3/+U5s3b9azzz6rrl27SvrhhveynsZlN8NnZGRoypQpOn/+vPr27etcJ/cxVQw1sN6uXbucHRMKCgqUlZWlv/3tb4qPj9fkyZPVt29fHT58WL6+vs6aSKV1ycvL00033XTZ01yoQcVQA+tRgxsLwRKmy8zM1BNPPKGYmBjFx8erXr16yszM1G9+8xutXLnS+cZR9saQm5srh8OhWrVqKTMzU5MnT1Z+fr7WrVvH2HAuogbW++677/TUU0/pvvvu04QJE+Tr66sjR45o6tSpmjlzpsaMGaPf/e53Cg8PV0lJidLT02W329W0aVMdPXpUGRkZuvXWW63ejSqNGliPGtx4CJYw3cmTJ1VQUKD+/furXr16KikpUcOGDdWnTx+tWrVKWVlZCgwMdI5DVlRUpJKSEmVmZmrq1KnKzc3lSSKVRA2sl5+fr5ycHN15551q2LChJKlNmzZauXKlxo8fr+XLl8swDP3hD39QYWGh1q5dq6SkJLVu3dp5W8LUqVMlMeizq6iB9ajBjYdPC5ju4sWLOnnypIqKiiT9cFasRYsWunDhghwOR7nBbevXr69bbrlFb731llq1akWgMQE1sJ7NZtOZM2f03XffOacVFRXJ399fixYt0pgxY7RixQo1a9ZMPXr00D333KN69eopKytLQUFBGjp0KGeLK4kaWI8a3Hj4xIDpAgIC1LBhQ+3atUtt27a97N4Yu91e7v9eXl5q1aqV8vPznc/bJdBUDjWwlmEYCggI0G233aZ169bp1ltvVfPmzZ1h3d/fX0lJSfrNb36jpUuXqkePHmrbtq3atm1bbj18mLqOGliPGtyYGG4IpmvRooX+/Oc/q3379qpVq5bzbFnZGbLi4mLnvJmZmdq+fbuef/55/b//9/8INCahBtay2Wzy8fFR//79tXPnTr377rv69ttvJZWG+LIzNs8995x27dqlLVu2XHE9fJi6jhpYjxrcmPjkgKnK7oHp3bu3c1rZc13PnTvnfMZ0SUmJTp06pYkTJ6pJkyaKioqSh4eHSkpKCDSVRA2sV1aDvn376siRI1q4cKF8fHw0ZMgQNWjQwHnGODAwUJ6ensrPz7e4xdUPNbAeNbgx8ekBU13pxuqys2SZmZny9fVV3bp1dfLkSU2ZMkUXLlzQvHnzLjujBtdRA+vZbDZnmJ84caLy8vKUmJio8+fPKzY2VhEREZJKe+PXrVv3soHpUXnUwHrU4MZEsMR1U/YGc+TIEU2bNk15eXl6//336SRyHVGD6+fSgP7YY4/Jz89PSUlJ2rVrl+666y7ZbDZt3LhRDRo00J133mldQ6sxamA9anDjsRmXPrwTuIZef/11zZs3TzfffLMk6R//+AeB5jqjBtb68MMPtXHjRn3yySdq1KiRQkJC9Morr8hut9NB4TqhBtajBtUbwRLXzUcffaRJkyapffv2evvtt+kkYgFqYI2yM8Vl/z537pzsdrv8/Pxks9mowXVADaxHDW4MBEu47NI3iauRl5ent99+W6NGjSLQmIQaWK+iNfjxIM8M+lx51MB61ABlCJa4KocPH9bBgwe1f/9+RUREqH379rrppptcXh+BpuKogfXS09O1b98+ZWZmqmPHjgoLC1OdOnW4fHcdUQPrUQP8HIIlftHOnTv1yCOPKCwsTCdOnJC3t7fOnj2rl19+WV27duWN5DqgBtbbuXOn4uLiVKtWLZ09e1Z+fn4KDg7W888/r+bNm1vdvBsCNbAeNcAvYVwR/Kxjx45p8uTJiomJ0UsvvaSNGzfq6aefVmhoqCZPnqx//vOf5Qbb/rGSkpLr2NrqiRpYLysrSzNnztSvf/1rLVu2TF9++aUmTJigwsJC/fa3v9X27dsllV7OuxJqUHnUwHrUAFeDYImfdfDgQRUUFOhXv/qVGjZsKEmKiorSvffeq/Pnz+upp57Sxx9/LKn8m8aZM2cklQ414XA4rn/DqxFqYL2zZ8/qu+++U+/evZ1nZX7/+9/rySefVKtWrRQXF6ddu3bJZrOVO9bUwDzUwHrUAFeDYImflZ6erpycHDVr1kySnE9GCA0N1YABA3T33Xfr8ccf18GDB+Xh4SHDMJSdna2pU6fqwQcflFT6OC7uuHAdNbDehQsXyt0/VlhYKJvNpk6dOmn69Olq0aKFxo4dq6NHjzqPNTUwFzWwHjXA1SBY4md17txZnp6eWr58uSTJx8dHkrR8+XJ5enrqwQcfVNOmTTV79mzl5uY6/1Lt1q2bfHx89Nxzz0m68tNgcHWowfX34w++li1bys/PT8nJyZIkb29vFRUVSZIiIyP18MMPq3Hjxnr++eepgUmogfWoAVxBsEQ5ubm5OnPmjE6fPq2LFy8qIiLCeT/NhAkT9OKLL+qBBx7Q8ePH9dhjj6lly5a6++67deDAAeXl5UmS6tWrp9GjR2v06NEKCAiwdoeqIGpgvR9frqtVq5YefvhhbdmyRXPnzpUk2e1254dq586d9atf/Up79+5Vdna2JGpQWdTAetQArmCsETh9+eWXeuGFF3Ty5EllZ2erd+/eiouL0/Tp09W0aVOtXbtWx44dU5MmTZSUlKTatWtLku677z4tXrxYhw8fdt4DKElhYWEKDAyUxBhlV4saWG/v3r166623dPz4cXl7e+uRRx5Ru3btdNtttyk2NlbvvPOOatasqQkTJshut6uwsFDe3t4aMGCAXnnlFaWmpqpJkybO9VGDiqMG1qMGcBXBEpKkPXv26KGHHtLdd9+t2NhYnTp1Sn//+9917tw5vfLKK5o4caJGjhwpLy8v1ahRo9yyhw4dUpMmTdSoUaNy0/39/eXv7y+JSx9XgxpY78svv9TIkSPVrVs3NWzYUOnp6XrooYe0cuVKhYaGatiwYcrLy9PSpUuVn5+vadOmydvbW5KUmpqqhg0bOj88y1CDiqEG1qMGqAyCJVRYWKikpCR16dJFf/rTn+Tn5yep9BLHvHnzdPz4cbVu3Vq+vr6SSt90Pv/8c3Xo0EEFBQV655131KxZMzVt2tTK3ajSqIH1Lly4oFdffVV33nmnnnnmGfn6+urQoUP64x//qI0bN+qhhx5Ss2bNNHnyZNWuXVtvvvmm9u/fr1/96lcqLCzUe++9p6ZNm6pt27ZW70qVRQ2sRw1QWdxjCZWUlCg9PV3h4eHy8/Nzjol4//33S5J2797tnLfsZu4VK1bowQcf1FNPPSUPDw8tWrRIHh4ejFPmImpgvcLCQmVkZKhVq1bOAF+nTh3VrFlTWVlZmjNnjpYuXarz58/rz3/+s1599VVlZmbqpZde0uuvv666devqzTffZEiVSqAG1qMGqCzOWEI1a9aU3W7X/v37Jcn5mL/69evLbrc7O4RIpZcw2rVrp48++kj79++Xr6+vWrRoIQ8PDx4RWAnUwHq+vr4qKipy1kCS9u/fr8zMTH3++efKz89XTk6O3n77bb344ouKjo5Wz549dfLkSdntdjVq1Eg2m40aVAI1sB41QGVxxvIGZxiGDMNQTEyMfHx8dOTIEUmlz5EuKCiQp6fnZffzXbhwQX5+foqKilKrVq2cZ8l4E3ENNXAPhmHo/vvvV+/evVVYWChJWr16te68804tXrxY//rXv/TSSy/J29tbq1evVlFRkby9vRUSEqLGjRvLZrNRg0qiBtajBqgsnhUOSVJeXp5OnTpV7lmvmZmZ6tWrl2bPnq1BgwZJkk6cOKGFCxdq6NChat26tVXNrZaogXXKeqkWFRXJbreXm/7dd9+pbt26zg4Hs2bN0qeffqoPPvhAtWrVsqrJ1Q41sB41gBn4kwIyDEO+vr7OQFP25lL2N0fNmjUlSRkZGZo8ebIuXLigiIgIy9pbHVEDa5V9WF76YVp2Ka9evXrO/3t6esrhcCg4OJgPU5NRA+tRA5iBS+G4bOiHsv+XPd81ICBA3377raZMmaK8vDytWbNGXl5e3JhtImrgfsou5ZXVwMvLS8ePH9dXX32l8PBwK5t2w6AG1qMGqCjOWN5gSkpK5OHxw98TPzdQbdnTFI4fP67XXntN58+f17p162S327kxuxKogfWutgaHDh1SUlKSzpw5o0aNGumrr76SzWbT448//rPL4ZdRA+tRA1wLnLG8gRiGIQ8PD128eFE7duyQpHKXW3+sadOmqlOnjp5++mllZ2cTaExADaxXkRoEBgbKMAxlZWUpIyNDnTp10urVq+Xl5aXi4mI+TF1EDaxHDXCt0HnnBuFwOJz3xTzzzDPau3evpkyZojvuuEPSlf/izMvL069//Wv5+flp7dq1zjcRAo1rqIH1KlKDsn87HA7nh23ZcS9bDyqOGliPGuBa4tPpBlD24s/Pz9cHH3ygnTt36vDhw1qwYIFsNpu6du3q/Ev10jcTX19f/f3vf1dgYKA8PT0JNJVADaxX0RqU1eHHH5yGYfBh6iJqYD1qgGuNM5Y3iPz8fA0YMECBgYG66aabVLt2ba1bt05t27bVmDFj1KNHD0nl/1K99P4b/jKtPGpgPVdqAHNRA+tRA1xLnPq4QSxZskQeHh568cUX1aRJE3l4eOiuu+7S/PnzlZSUJC8vr8v+Ur30pm4CTeVRA+u5UgOYixpYjxrgWiJY3gAMw9Dhw4fVsGFDBQcHyzAMlZSUqFevXpKkRx55RK+++qoMw9Add9zBm8g1QA2sRw2sRw2sRw1wrdErvBoqLi4u93+bzabatWvr2LFjunDhgvOv0LI3kwEDBujLL7/UG2+8oZ07d1rU6uqFGliPGliPGliPGuB6I1hWQ15eXsrPz1dCQoKysrIkSR07dtT58+eVnJys/Px8eXp6Onv65ebmqlWrVtq7d69SUlIk6SeHv8HVoQbWowbWowbWowa43giW1dS///1vLVq0SPPnz1dubq4GDhyo9u3ba8mSJUpOTlZBQYFsNpuOHTumw4cPa9KkSZo4caLWrl2rkydPcvnDBNTAetTAetTAetQA1xO9wqupwsJCrVu3Ts8884zuu+8+Pf/887Lb7Ro/frx27dqlmjVrqmnTpsrIyJCvr6/Wrl2rjRs3av78+Vq1apXq1Klj9S5UedTAetTAetTAetQA1xOdd6qBsmFoLh2axtvbW/3791dJSYlmz54twzD0/PPPa9GiRfrHP/6hnTt36vz587rttts0ceJESdKmTZt000030fvYBdTAetTAetTAetQAljNQLVy4cMGYNm2a8fnnn5ebfvHiRePvf/+70aZNG+Oxxx4zsrOznT/Lz8839u3bZ2zZssWYNm2a0blzZyM1NfV6N73aoAbWowbWowbWowawEmcsq7BLB8z+z3/+o82bN+vYsWOy2+3q1KmTpNK/VPv27ausrCwlJiaqTp06GjlypG666SYdOXJEr776qv73v/+pefPmeuuttxQREWHlLlU51MB61MB61MB61ABuw+pki4o5fPiw8Z///OeKP1uzZo1xzz33GIMHDzZ27NhhGIZhlJSUGIZhGP/973+Ntm3bGi1atDAWLFjgXOabb74xjh07Vu4vV/w8amA9amA9amA9agB3ROedKuTYsWPq06ePatWqpbCwMI0ePVotW7ZU06ZNnfOsXr1aS5cuVZ06dTR9+nRFRUVJkj755BNt2bJFv/3tbxUeHs7zpl1EDaxHDaxHDaxHDeCu+G2qQs6cOSPDMHTnnXfq4sWLmjp1qho0aKA//OEP6tixo9q1a6dBgwbJ29tbiYmJevrppxUXFyd/f3+99tprCgoKUsuWLSWVDprLm0nFUQPrUQPrUQPrUQO4K85YVjHPPPOMtm7dqn/84x/aunWrPvroI73//vsKCgpS165dNXjwYHXs2FGffPKJ5s+fr/3796t27doKDQ3VO++8I7vdzrNfK4kaWI8aWI8aWI8awB0RLKuIsqEjPv/8c02fPl0TJkzQkCFDJJVeEpk4caJSU1Pl6+urW265RQ8//LDy8vLUoEEDZWZm6u6775anpyd/mVYCNbAeNbAeNbAeNYA748k7VUTZeGS33367QkND9e677zp/9tVXX+mbb77RlClTNGDAAGVnZ2vUqFGaNGmSzpw5o3vuuUeenp5yOBy8iVQCNbAeNbAeNbAeNYBbu/79heCq4uJiwzAM44svvjDatm1rfPzxx8a6deuMFi1aGPPmzXPOl5WVZSQnJxtxcXFGUVGRVc2tlqiB9aiB9aiB9agB3BWXwqugzMxMjRs3Tjk5OcrMzNTo0aM1ZswY+fj4XHF+LneYjxpYjxpYjxpYjxrA3XApvApq2LChhg8froyMDA0aNEiPPPJIuTeRsr8Vyr7zJmI+amA9amA9amA9agB3Q7Csojp16qTIyEgdP35cubm55X5W1sOPnn7XFjWwHjWwHjWwHjWAOyFYVlHBwcGKiorSjh07dPToUUmlPQVx/VAD61ED61ED61EDuBOCZRVUdklj3LhxqlOnjhISEiT90FMQ1x41sB41sB41sB41gLvhN68KKrukUatWLTVv3ly5ubmiD9b1RQ2sRw2sRw2sRw3gbugVXsWdOnVKDRo0kKenp3PQXFxf1MB61MB61MB61ADugGBZTTgcDnl6elrdjBsaNbAeNbAeNbAeNYCVCJYAAAAwBefJAQAAYAqCJQAAAExBsAQAAIApCJYAAAAwBcESAAAApiBYAgAAwBQESwAAAJiCYAkAAABTECwBAABgCoIlAAAATPH/ASrXt5oyu8kHAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 800x575 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x575 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(1, len(contracts)):\n",
    "    last_index_end = indexes[i-1][0][-5]\n",
    "    this_index_start = indexes[i][0][5]\n",
    "    tmp_df = train_df.iloc[last_index_end:this_index_start].set_index('date')\n",
    "    tmp_df.index = pd.to_datetime(tmp_df.index)\n",
    "    mpf.plot(tmp_df, type='candle')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "264bc23e",
   "metadata": {},
   "source": [
    "# Get All Features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "56a11155",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-04-27T03:13:27.000289Z",
     "start_time": "2023-04-27T03:06:45.453271Z"
    },
    "code_folding": [
     4,
     9,
     17
    ]
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████| 6433/6433 [06:27<00:00, 16.60it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Error:  underlaying_symbols\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "((6433, 144),\n",
       " date              object\n",
       " tic               object\n",
       " open             float64\n",
       " high             float64\n",
       " low              float64\n",
       " close            float64\n",
       " volume           float64\n",
       " open_oi          float64\n",
       " close_oi         float64\n",
       " alpha_191_1      float64\n",
       " alpha_191_2      float64\n",
       " alpha_191_5      float64\n",
       " alpha_191_6      float64\n",
       " alpha_191_7      float64\n",
       " alpha_191_8      float64\n",
       " alpha_191_9      float64\n",
       " alpha_191_11     float64\n",
       " alpha_191_12     float64\n",
       " alpha_191_13     float64\n",
       " alpha_191_14     float64\n",
       " alpha_191_15     float64\n",
       " alpha_191_16     float64\n",
       " alpha_191_17     float64\n",
       " alpha_191_18     float64\n",
       " alpha_191_20     float64\n",
       " alpha_191_22     float64\n",
       " alpha_191_24     float64\n",
       " alpha_191_25     float64\n",
       " alpha_191_26     float64\n",
       " alpha_191_28     float64\n",
       " alpha_191_29     float64\n",
       " alpha_191_31     float64\n",
       " alpha_191_32     float64\n",
       " alpha_191_33     float64\n",
       " alpha_191_34     float64\n",
       " alpha_191_35     float64\n",
       " alpha_191_36     float64\n",
       " alpha_191_37     float64\n",
       " alpha_191_39     float64\n",
       " alpha_191_41     float64\n",
       " alpha_191_42     float64\n",
       " alpha_191_44     float64\n",
       " alpha_191_45     float64\n",
       " alpha_191_46     float64\n",
       " alpha_191_47     float64\n",
       " alpha_191_48     float64\n",
       " alpha_191_52     float64\n",
       " alpha_191_54     float64\n",
       " alpha_191_57     float64\n",
       " alpha_191_60     float64\n",
       " alpha_191_61     float64\n",
       " alpha_191_62     float64\n",
       " alpha_191_63     float64\n",
       " alpha_191_64     float64\n",
       " alpha_191_65     float64\n",
       " alpha_191_66     float64\n",
       " alpha_191_67     float64\n",
       " alpha_191_68     float64\n",
       " alpha_191_71     float64\n",
       " alpha_191_72     float64\n",
       " alpha_191_73     float64\n",
       " alpha_191_74     float64\n",
       " alpha_191_76     float64\n",
       " alpha_191_77     float64\n",
       " alpha_191_78     float64\n",
       " alpha_191_79     float64\n",
       " alpha_191_80     float64\n",
       " alpha_191_81     float64\n",
       " alpha_191_82     float64\n",
       " alpha_191_83     float64\n",
       " alpha_191_85     float64\n",
       " alpha_191_87     float64\n",
       " alpha_191_88     float64\n",
       " alpha_191_89     float64\n",
       " alpha_191_90     float64\n",
       " alpha_191_91     float64\n",
       " alpha_191_92     float64\n",
       " alpha_191_95     float64\n",
       " alpha_191_96     float64\n",
       " alpha_191_97     float64\n",
       " alpha_191_99     float64\n",
       " alpha_191_100    float64\n",
       " alpha_191_70     float64\n",
       " alpha_191_102    float64\n",
       " alpha_191_104    float64\n",
       " alpha_191_105    float64\n",
       " alpha_191_106    float64\n",
       " alpha_191_107    float64\n",
       " alpha_191_108    float64\n",
       " alpha_191_109    float64\n",
       " alpha_191_110    float64\n",
       " alpha_191_111    float64\n",
       " alpha_191_113    float64\n",
       " alpha_191_114    float64\n",
       " alpha_191_115    float64\n",
       " alpha_191_117    float64\n",
       " alpha_191_118    float64\n",
       " alpha_191_119    float64\n",
       " alpha_191_120    float64\n",
       " alpha_191_121    float64\n",
       " alpha_191_122    float64\n",
       " alpha_191_124    float64\n",
       " alpha_191_125    float64\n",
       " alpha_191_126    float64\n",
       " alpha_191_130    float64\n",
       " alpha_191_131    float64\n",
       " alpha_191_134    float64\n",
       " alpha_191_135    float64\n",
       " alpha_191_136    float64\n",
       " alpha_191_138    float64\n",
       " alpha_191_139    float64\n",
       " alpha_191_140    float64\n",
       " alpha_191_141    float64\n",
       " alpha_191_142    float64\n",
       " alpha_191_145    float64\n",
       " alpha_191_146    float64\n",
       " alpha_191_150    float64\n",
       " alpha_191_132    float64\n",
       " alpha_191_151    float64\n",
       " alpha_191_152    float64\n",
       " alpha_191_153    float64\n",
       " alpha_191_155    float64\n",
       " alpha_191_156    float64\n",
       " alpha_191_157    float64\n",
       " alpha_191_158    float64\n",
       " alpha_191_159    float64\n",
       " alpha_191_161    float64\n",
       " alpha_191_162    float64\n",
       " alpha_191_163    float64\n",
       " alpha_191_168    float64\n",
       " alpha_191_169    float64\n",
       " alpha_191_170    float64\n",
       " alpha_191_171    float64\n",
       " alpha_191_173    float64\n",
       " alpha_191_175    float64\n",
       " alpha_191_176    float64\n",
       " alpha_191_178    float64\n",
       " alpha_191_179    float64\n",
       " alpha_191_181    float64\n",
       " alpha_191_184    float64\n",
       " alpha_191_185    float64\n",
       " alpha_191_188    float64\n",
       " alpha_191_189    float64\n",
       " alpha_191_191    float64\n",
       " dtype: object,\n",
       "                   date tic       open       high        low      close  \\\n",
       " 0  2022-08-26 14:30:00  RB 4093.00000 4100.00000 4093.00000 4099.00000   \n",
       " 1  2022-08-26 14:35:00  RB 4099.00000 4104.00000 4096.00000 4100.00000   \n",
       " 2  2022-08-26 14:40:00  RB 4100.00000 4109.00000 4100.00000 4106.00000   \n",
       " 3  2022-08-26 14:45:00  RB 4106.00000 4112.00000 4093.00000 4095.00000   \n",
       " 4  2022-08-26 14:50:00  RB 4095.00000 4097.00000 4088.00000 4097.00000   \n",
       " \n",
       "        volume       open_oi      close_oi  alpha_191_1  alpha_191_2  \\\n",
       " 0 18553.00000 1142068.00000 1142878.00000     -0.60450     -0.00000   \n",
       " 1 27010.00000 1142878.00000 1141798.00000     -0.56644      0.71429   \n",
       " 2 24026.00000 1141798.00000 1145783.00000     -0.16547     -0.33333   \n",
       " 3 45856.00000 1145783.00000 1137845.00000      0.29970      1.12281   \n",
       " 4 36535.00000 1137845.00000 1118974.00000      0.26372     -1.78947   \n",
       " \n",
       "    alpha_191_5  alpha_191_6  alpha_191_7  alpha_191_8  alpha_191_9  \\\n",
       " 0     -0.83918     -0.22314      0.89233      0.32645      0.00040   \n",
       " 1     -0.93917     -0.72314      0.94190      0.26446      0.00058   \n",
       " 2     -0.93917     -0.72314      0.92538      0.09917      0.00090   \n",
       " 3     -0.94324     -0.72314      0.96669      0.41322      0.00040   \n",
       " 4     -0.94324     -0.72314      0.82623      0.66942     -0.00042   \n",
       " \n",
       "    alpha_191_11  alpha_191_12  alpha_191_13  alpha_191_14  alpha_191_15  \\\n",
       " 0   26203.64286      -0.20203      -2.50150       4.00000       0.00000   \n",
       " 1   32231.50000      -0.34107      -0.00195       7.00000       0.00000   \n",
       " 2   41644.88095      -0.33455      -1.50247      12.00000       0.00000   \n",
       " 3    -980.22431      -0.47359       7.48900       2.00000       0.00000   \n",
       " 4   27792.27569      -0.18900      -4.50247       4.00000       0.00000   \n",
       " \n",
       "    alpha_191_16  alpha_191_17  alpha_191_18  alpha_191_20  alpha_191_22  \\\n",
       " 0      -1.00000       0.06452       1.00098       0.04882      -0.00004   \n",
       " 1      -0.96694       0.00826       1.00171       0.12210       0.00008   \n",
       " 2      -1.00000       0.00027       1.00293       0.31762       0.00027   \n",
       " 3      -1.00000       0.25402       1.00049       0.02443       0.00010   \n",
       " 4      -1.00000       0.06452       1.00098       0.09773      -0.00003   \n",
       " \n",
       "    alpha_191_24  alpha_191_25  alpha_191_26  alpha_191_28  alpha_191_29  \\\n",
       " 0       1.88100           NaN           NaN    -885.70762       9.05687   \n",
       " 1       2.90480           NaN           NaN    -792.29149      32.97924   \n",
       " 2       4.72384           NaN           NaN    -677.01887      76.31029   \n",
       " 3       4.17907           NaN           NaN    -590.58457      11.20078   \n",
       " 4       4.14326           NaN           NaN    -385.83657      35.70486   \n",
       " \n",
       "    alpha_191_31  alpha_191_32  alpha_191_33  alpha_191_34  alpha_191_35  \\\n",
       " 0       0.14659      -1.80488           NaN       0.99854      -0.55645   \n",
       " 1       0.14248      -2.16260           NaN       0.99858      -0.77419   \n",
       " 2       0.25026      -2.44715           NaN       0.99750      -0.76613   \n",
       " 3      -0.02238      -2.11382           NaN       1.00022      -0.87097   \n",
       " 4       0.02034      -1.65854           NaN       0.99980      -0.45968   \n",
       " \n",
       "    alpha_191_36  alpha_191_37  alpha_191_39  alpha_191_41  alpha_191_42  \\\n",
       " 0       0.96639      -0.48182      -0.03220      -0.38934      -0.06120   \n",
       " 1       0.97479      -0.49091      -0.18667      -0.81967      -0.13718   \n",
       " 2       1.00000      -0.66364      -0.28830      -0.91803      -0.38616   \n",
       " 3       0.96639      -0.49091      -0.12976      -0.38115      -0.55910   \n",
       " 4       0.90756      -0.37273       0.05723      -0.38115      -0.46255   \n",
       " \n",
       "    alpha_191_44  alpha_191_45  alpha_191_46  alpha_191_47  alpha_191_48  \\\n",
       " 0      14.00000           NaN       0.99866      35.60726      -0.10288   \n",
       " 1      15.00000           NaN       0.99871      34.12003      -0.23808   \n",
       " 2      19.00000           NaN       0.99778      31.77817      -0.30610   \n",
       " 3      15.00000           NaN       1.00041      35.51221      -0.27579   \n",
       " 4      10.00000           NaN       0.99992      38.23308      -0.31933   \n",
       " \n",
       "    alpha_191_52  alpha_191_54  alpha_191_57  alpha_191_60  alpha_191_61  \\\n",
       " 0     132.38866      -0.85345      71.68610   38406.37619      -0.86239   \n",
       " 1     138.93443      -0.51724      73.71666   31330.37619      -0.87156   \n",
       " 2     141.90871      -0.87931      78.12994   37006.24286      -0.88991   \n",
       " 3     121.01449      -0.03448      63.62509    7188.13759      -0.89908   \n",
       " 4     102.95082      -0.60345      56.51929   55021.13759      -0.91743   \n",
       " \n",
       "    alpha_191_62  alpha_191_63  alpha_191_64  alpha_191_65  alpha_191_66  \\\n",
       " 0      -0.87745      71.31234      -0.78125       0.99890       0.10990   \n",
       " 1      -0.97633      73.36048      -0.81967       0.99886       0.11395   \n",
       " 2      -0.93843      82.40500      -0.95082       0.99793       0.20744   \n",
       " 3      -0.91333      47.17123      -0.78689       1.00065      -0.06508   \n",
       " 4      -0.35267      51.67898      -0.78125       1.00033      -0.03253   \n",
       " \n",
       "    alpha_191_67  alpha_191_68  alpha_191_71  alpha_191_72  alpha_191_73  \\\n",
       " 0      59.37147       0.00027       0.18229      36.04333      -4.13821   \n",
       " 1      59.78433       0.00037       0.19448      35.12313      -4.07317   \n",
       " 2      62.67685       0.00055       0.32579      33.64209      -4.04065   \n",
       " 3      55.58965       0.00036       0.05498      35.75657      -4.26829   \n",
       " 4      56.08239      -0.00001       0.10180      37.37324      -4.75610   \n",
       " \n",
       "    alpha_191_74  alpha_191_76  alpha_191_77  alpha_191_78  alpha_191_79  \\\n",
       " 0       1.50765       0.78738       0.10400      89.73294      64.54475   \n",
       " 1       1.47514       0.79173       0.11200     107.52928      65.60996   \n",
       " 2       1.51626       0.75093       0.10400     138.40304      71.26352   \n",
       " 3       1.49071       0.73413       0.41600      62.83142      53.63246   \n",
       " 4       1.52268       0.79176       0.18400     -26.41108      55.79945   \n",
       " \n",
       "    alpha_191_80  alpha_191_81  alpha_191_82  alpha_191_83  alpha_191_85  \\\n",
       " 0      31.90899   12858.67204      36.85575      -0.86777      50.00000   \n",
       " 1     174.68728   14206.48986      36.13418      -0.99174     120.00000   \n",
       " 2     274.06196   15141.52811      34.90384      -0.99174      19.00000   \n",
       " 3     343.05314   18066.69779      36.41373      -0.96694     120.00000   \n",
       " 4     321.00714   19825.57719      37.59679      -0.47107      85.50000   \n",
       " \n",
       "    alpha_191_87  alpha_191_88  alpha_191_89  alpha_191_90  alpha_191_91  \\\n",
       " 0      -7.50000       0.09768      -0.46202      -0.95041      -0.25815   \n",
       " 1      -7.54545       0.09766      -0.02909      -0.96694      -0.35649   \n",
       " 2      -7.73554       0.26862       0.92354      -1.00000      -0.41795   \n",
       " 3      -7.70248       0.07331      -0.01989      -0.86777      -0.34420   \n",
       " 4      -5.62810       0.31832      -0.44249      -0.36364      -0.16595   \n",
       " \n",
       "    alpha_191_92  alpha_191_95  alpha_191_96  alpha_191_97  alpha_191_99  \\\n",
       " 0      -8.00000    4363.14490      68.16023    3673.77679      -0.83471   \n",
       " 1      -8.00000    5072.65498      70.01237    6083.69270      -0.96694   \n",
       " 2      -8.00000    5593.83768      72.71823    6926.34825      -0.99174   \n",
       " 3      -8.00000    9042.13329      69.68718   11843.19057      -0.91736   \n",
       " 4      -8.00000   10163.31720      65.29789   13137.91015      -0.47934   \n",
       " \n",
       "    alpha_191_100  alpha_191_70  alpha_191_102  alpha_191_104  alpha_191_105  \\\n",
       " 0     4363.14490    4335.06867       64.74394        0.00428       -0.14967   \n",
       " 1     5072.65498    7807.09945       74.61821       -0.01354       -0.39725   \n",
       " 2     5593.83768    8593.64996       66.70759        0.00426       -0.55338   \n",
       " 3     9042.13329   13582.15165       82.75614       -0.00608       -0.66473   \n",
       " 4    10163.31720   13130.55269       66.36468        0.04960       -0.72153   \n",
       " \n",
       "    alpha_191_106  alpha_191_107  alpha_191_108  alpha_191_109  alpha_191_110  \\\n",
       " 0        4.00000       -0.31165       -0.93600        0.99192       79.48718   \n",
       " 1        4.00000       -0.31738       -0.96400        1.02253       82.05128   \n",
       " 2       11.00000       -0.12239       -0.98800        1.06352       97.29730   \n",
       " 3        3.00000       -0.20535       -1.00000        1.29389       92.59259   \n",
       " 4       13.00000       -0.00266       -0.81200        1.20825       97.40260   \n",
       " \n",
       "    alpha_191_111  alpha_191_113  alpha_191_114  alpha_191_115  alpha_191_117  \\\n",
       " 0    -4090.59717       -0.63852        0.00000        0.72120     9720.00000   \n",
       " 1     -329.65093       -0.66851        0.00000        0.84517     6510.00000   \n",
       " 2    -1309.32586        0.69608        0.00000        1.00000    11310.00000   \n",
       " 3    12475.99145        0.69118        0.00000        1.00000       -0.00000   \n",
       " 4    -6162.36601        0.68288        0.00000        0.82004     6417.00000   \n",
       " \n",
       "    alpha_191_118  alpha_191_119  alpha_191_120  alpha_191_121  alpha_191_122  \\\n",
       " 0       79.48718       -0.03725        0.51852       -0.97200        0.00002   \n",
       " 1       82.05128       -0.17156        0.51639       -0.97600        0.00002   \n",
       " 2       97.29730       -0.11259        0.50400       -1.00000        0.00002   \n",
       " 3       92.59259       -0.16344        0.58333       -0.74650        0.00002   \n",
       " 4       97.40260       -0.26456        0.55022       -0.76858        0.00002   \n",
       " \n",
       "    alpha_191_124  alpha_191_125  alpha_191_126  alpha_191_130  alpha_191_131  \\\n",
       " 0        0.00000        0.15045     4097.33333        0.96315        0.46135   \n",
       " 1        0.00000        0.10832     4100.00000        0.93966        0.17230   \n",
       " 2        0.00000        0.07822     4105.00000        0.93966        0.39270   \n",
       " 3        0.00000        0.05004     4100.00000        0.98082        0.00000   \n",
       " 4        0.00000        0.01404     4094.00000        0.99691        0.01633   \n",
       " \n",
       "    alpha_191_134  alpha_191_135  alpha_191_136  alpha_191_138  alpha_191_139  \\\n",
       " 0       45.37295        1.00307       -0.13894        5.52459       -0.18475   \n",
       " 1       92.54528        1.00296       -0.37646        4.43443       -0.61557   \n",
       " 2      111.69415        1.00287       -0.57170        2.19672       -0.72058   \n",
       " 3       22.40704        1.00286       -0.02229        0.20492       -0.89888   \n",
       " 4       26.77211        1.00274       -0.41517        0.36066       -0.77285   \n",
       " \n",
       "    alpha_191_140  alpha_191_141  alpha_191_142  alpha_191_145  alpha_191_146  \\\n",
       " 0        0.54400       -0.40777       -0.64223      -28.76837       -6.53767   \n",
       " 1        0.53600       -0.42718       -0.21306      -17.47967       -0.39126   \n",
       " 2        0.46400       -0.63107       -0.47663       -4.25877        2.80107   \n",
       " 3        0.48800       -0.73786       -0.01687       12.18644        2.83704   \n",
       " 4        0.19200       -0.59223       -0.37802       21.13633        1.53424   \n",
       " \n",
       "     alpha_191_150  alpha_191_132  alpha_191_151  alpha_191_152  alpha_191_153  \\\n",
       " 0  76017825.33333    12974.70000       12.03946        0.00052     4093.51042   \n",
       " 1 110741000.00000    13263.80000       11.61458        0.00038     4094.71875   \n",
       " 2  98626730.00000    13881.90000       11.60675        0.00024     4096.89583   \n",
       " 3 188009600.00000    15217.10000       11.19016        0.00012     4096.66667   \n",
       " 4 149574290.00000    15994.75000       11.24400        0.00002     4096.66667   \n",
       " \n",
       "    alpha_191_155  alpha_191_156  alpha_191_157  alpha_191_158  alpha_191_159  \\\n",
       " 0      277.32037       -0.48780        2.54622        0.00171   -60983.13769   \n",
       " 1     1234.98624       -0.62602        4.42437        0.00195   -60463.01495   \n",
       " 2     1588.46793       -0.83740        4.20168        0.00219   -58364.91617   \n",
       " 3     3127.07653       -0.71951        2.19748        0.00464   -48680.07245   \n",
       " 4     3332.05102       -0.83740        3.54622        0.00220   -47457.21300   \n",
       " \n",
       "    alpha_191_161  alpha_191_162  alpha_191_163  alpha_191_168  alpha_191_169  \\\n",
       " 0        6.58333        1.11289        0.29245       -1.42994       -0.02291   \n",
       " 1        6.58333        1.12285        0.36792       -2.03637       -0.05585   \n",
       " 2        6.91667        1.21680        0.09434       -1.73074       -0.07849   \n",
       " 3        7.83333        0.86490        0.97170       -3.01345       -0.09006   \n",
       " 4        8.16667        0.91720        0.50472       -2.28419       -0.09873   \n",
       " \n",
       "    alpha_191_170  alpha_191_171  alpha_191_173  alpha_191_175  alpha_191_176  \\\n",
       " 0       -0.57385       -5.95622     4107.99915        6.66667        0.80783   \n",
       " 1       -0.66761       -0.99878     4109.82784        6.83333        0.75798   \n",
       " 2       -0.86001       -1.98543     4113.58660        7.16667        0.78438   \n",
       " 3       -0.05791       -0.11924     4111.73888        9.66667        0.11551   \n",
       " 4       -0.57159            inf     4111.04504        9.83333       -0.15960   \n",
       " \n",
       "    alpha_191_178  alpha_191_179  alpha_191_181  alpha_191_184  alpha_191_185  \\\n",
       " 0       27.19717        0.42055       -0.06065            NaN        0.28000   \n",
       " 1        6.58941        0.54590       -0.07742            NaN        0.95200   \n",
       " 2       35.16000        0.64603       -0.09882            NaN        0.28800   \n",
       " 3     -122.84851        0.27793       -0.11512            NaN        0.06400   \n",
       " 4       17.84371        0.08361       -0.11473            NaN        0.76800   \n",
       " \n",
       "    alpha_191_188  alpha_191_189  alpha_191_191  \n",
       " 0        3.40753        1.63889       -2.45691  \n",
       " 1       14.39863        2.08333       -0.23250  \n",
       " 2       22.31613        3.33333       -1.24948  \n",
       " 3      100.53377        3.75000        7.77465  \n",
       " 4       -4.13704        3.80556       -5.12675  )"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "window_size = 125\n",
    "individual_length = train_df[train_df.tic==tickers[0]].shape[0]\n",
    "columns = GET_alpha(train_df[:window_size+5]).columns.to_list()\n",
    "res_list = []\n",
    "def data_processing(i, tmp_df):\n",
    "    tmp_df_1 = tmp_df.iloc[i:i+window_size].copy()\n",
    "    res = GET_alpha(tmp_df_1)\n",
    "    clear_output()\n",
    "    return res.iloc[-1].to_numpy()\n",
    "for a in range(len(tickers)):\n",
    "    tmp_df = train_df[train_df.tic==tickers[a]]\n",
    "    res = Parallel(n_jobs=40)(delayed(data_processing)(i, tmp_df) for i in tqdm(range(individual_length-window_size)))\n",
    "    res_list.append(res)\n",
    "res_array = np.asarray(res_list)\n",
    "res_array = res_array.reshape(res_array.shape[0]*res_array.shape[1], res_array.shape[-1])\n",
    "res_df = pd.DataFrame(res_array)\n",
    "res_df.columns = columns\n",
    "for i in range(2, len(res_df.columns)):\n",
    "    try:\n",
    "        res_df[columns[i]] = res_df[columns[i]].values.astype('float64')\n",
    "    except:\n",
    "        print(\"Error: \", columns[i])\n",
    "        res_df = res_df.drop(columns=columns[i])\n",
    "\n",
    "res_df.shape, res_df.dtypes, res_df.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e91a3120",
   "metadata": {},
   "source": [
    "# Drop Features with Too Many Zeros\n",
    "## If data set contains too many zeros, the agent will be very reluctant to trade."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "c2bb9a66",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-04-27T03:16:16.452569Z",
     "start_time": "2023-04-27T03:16:16.412540Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(6,)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "columns_1 = res_df.columns\n",
    "zeros_1 = ((res_df==0).sum()>res_df.shape[0]*0.05)\n",
    "columns_to_drop_1 = np.where(zeros_1==True)[0]\n",
    "\n",
    "columns_to_drop_1.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "062a9aed",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-04-27T03:16:16.708736Z",
     "start_time": "2023-04-27T03:16:16.458572Z"
    },
    "code_folding": [
     0
    ]
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(date                0\n",
       " tic                 0\n",
       " open                0\n",
       " high                0\n",
       " low                 0\n",
       " close               0\n",
       " volume              0\n",
       " open_oi             0\n",
       " close_oi            0\n",
       " alpha_191_1         0\n",
       " alpha_191_2         0\n",
       " alpha_191_5       211\n",
       " alpha_191_6         0\n",
       " alpha_191_7         0\n",
       " alpha_191_8         0\n",
       " alpha_191_9         0\n",
       " alpha_191_11        0\n",
       " alpha_191_12        0\n",
       " alpha_191_13        0\n",
       " alpha_191_14        0\n",
       " alpha_191_16        0\n",
       " alpha_191_17        0\n",
       " alpha_191_18        0\n",
       " alpha_191_20        0\n",
       " alpha_191_22        0\n",
       " alpha_191_24        0\n",
       " alpha_191_25     6433\n",
       " alpha_191_26     6433\n",
       " alpha_191_28        0\n",
       " alpha_191_29        0\n",
       " alpha_191_31        0\n",
       " alpha_191_32      144\n",
       " alpha_191_33     6433\n",
       " alpha_191_34        0\n",
       " alpha_191_35        0\n",
       " alpha_191_36        0\n",
       " alpha_191_37        0\n",
       " alpha_191_39        0\n",
       " alpha_191_41        0\n",
       " alpha_191_42        0\n",
       " alpha_191_44        0\n",
       " alpha_191_45     6433\n",
       " alpha_191_46        0\n",
       " alpha_191_47        0\n",
       " alpha_191_48        0\n",
       " alpha_191_52        0\n",
       " alpha_191_54        0\n",
       " alpha_191_57        0\n",
       " alpha_191_60        0\n",
       " alpha_191_61        0\n",
       " alpha_191_62        2\n",
       " alpha_191_63        0\n",
       " alpha_191_64        0\n",
       " alpha_191_65        0\n",
       " alpha_191_66        0\n",
       " alpha_191_67        0\n",
       " alpha_191_68        0\n",
       " alpha_191_71        0\n",
       " alpha_191_72        0\n",
       " alpha_191_73        0\n",
       " alpha_191_74        0\n",
       " alpha_191_76        0\n",
       " alpha_191_77        0\n",
       " alpha_191_78        0\n",
       " alpha_191_79        0\n",
       " alpha_191_80        0\n",
       " alpha_191_81        0\n",
       " alpha_191_82        0\n",
       " alpha_191_83        0\n",
       " alpha_191_85        0\n",
       " alpha_191_87       48\n",
       " alpha_191_88        0\n",
       " alpha_191_89        0\n",
       " alpha_191_90        0\n",
       " alpha_191_91        0\n",
       " alpha_191_92        0\n",
       " alpha_191_95        0\n",
       " alpha_191_96        0\n",
       " alpha_191_97        0\n",
       " alpha_191_99        0\n",
       " alpha_191_100       0\n",
       " alpha_191_70        0\n",
       " alpha_191_102       0\n",
       " alpha_191_104       4\n",
       " alpha_191_105       0\n",
       " alpha_191_106       0\n",
       " alpha_191_107       0\n",
       " alpha_191_108       0\n",
       " alpha_191_109       0\n",
       " alpha_191_110       0\n",
       " alpha_191_111       0\n",
       " alpha_191_113     850\n",
       " alpha_191_115      24\n",
       " alpha_191_118       0\n",
       " alpha_191_119       0\n",
       " alpha_191_120       0\n",
       " alpha_191_121       1\n",
       " alpha_191_122       0\n",
       " alpha_191_125       0\n",
       " alpha_191_126       0\n",
       " alpha_191_130       0\n",
       " alpha_191_131       0\n",
       " alpha_191_134       0\n",
       " alpha_191_135       0\n",
       " alpha_191_136       0\n",
       " alpha_191_138       0\n",
       " alpha_191_139       0\n",
       " alpha_191_140       0\n",
       " alpha_191_141       0\n",
       " alpha_191_142       0\n",
       " alpha_191_145       0\n",
       " alpha_191_146       0\n",
       " alpha_191_150       0\n",
       " alpha_191_132       0\n",
       " alpha_191_151       0\n",
       " alpha_191_152       0\n",
       " alpha_191_153       0\n",
       " alpha_191_155       0\n",
       " alpha_191_156       0\n",
       " alpha_191_157       0\n",
       " alpha_191_158       0\n",
       " alpha_191_159       0\n",
       " alpha_191_161       0\n",
       " alpha_191_162       0\n",
       " alpha_191_163       0\n",
       " alpha_191_168       0\n",
       " alpha_191_169       0\n",
       " alpha_191_170       0\n",
       " alpha_191_173       0\n",
       " alpha_191_175       0\n",
       " alpha_191_176       0\n",
       " alpha_191_179       1\n",
       " alpha_191_181       0\n",
       " alpha_191_184    6433\n",
       " alpha_191_185       0\n",
       " alpha_191_188       0\n",
       " alpha_191_189       0\n",
       " alpha_191_191       0\n",
       " dtype: int64,\n",
       " (6433, 138),\n",
       " Index(['date', 'tic', 'open', 'high', 'low', 'close', 'volume', 'open_oi',\n",
       "        'close_oi', 'alpha_191_1',\n",
       "        ...\n",
       "        'alpha_191_173', 'alpha_191_175', 'alpha_191_176', 'alpha_191_179',\n",
       "        'alpha_191_181', 'alpha_191_184', 'alpha_191_185', 'alpha_191_188',\n",
       "        'alpha_191_189', 'alpha_191_191'],\n",
       "       dtype='object', length=138),\n",
       "                   date tic       open       high        low      close  \\\n",
       " 0  2022-08-26 14:30:00  RB 4093.00000 4100.00000 4093.00000 4099.00000   \n",
       " 1  2022-08-26 14:35:00  RB 4099.00000 4104.00000 4096.00000 4100.00000   \n",
       " 2  2022-08-26 14:40:00  RB 4100.00000 4109.00000 4100.00000 4106.00000   \n",
       " 3  2022-08-26 14:45:00  RB 4106.00000 4112.00000 4093.00000 4095.00000   \n",
       " 4  2022-08-26 14:50:00  RB 4095.00000 4097.00000 4088.00000 4097.00000   \n",
       " \n",
       "        volume       open_oi      close_oi  alpha_191_1  alpha_191_2  \\\n",
       " 0 18553.00000 1142068.00000 1142878.00000     -0.60450     -0.00000   \n",
       " 1 27010.00000 1142878.00000 1141798.00000     -0.56644      0.71429   \n",
       " 2 24026.00000 1141798.00000 1145783.00000     -0.16547     -0.33333   \n",
       " 3 45856.00000 1145783.00000 1137845.00000      0.29970      1.12281   \n",
       " 4 36535.00000 1137845.00000 1118974.00000      0.26372     -1.78947   \n",
       " \n",
       "    alpha_191_5  alpha_191_6  alpha_191_7  alpha_191_8  alpha_191_9  \\\n",
       " 0     -0.83918     -0.22314      0.89233      0.32645      0.00040   \n",
       " 1     -0.93917     -0.72314      0.94190      0.26446      0.00058   \n",
       " 2     -0.93917     -0.72314      0.92538      0.09917      0.00090   \n",
       " 3     -0.94324     -0.72314      0.96669      0.41322      0.00040   \n",
       " 4     -0.94324     -0.72314      0.82623      0.66942     -0.00042   \n",
       " \n",
       "    alpha_191_11  alpha_191_12  alpha_191_13  alpha_191_14  alpha_191_16  \\\n",
       " 0   26203.64286      -0.20203      -2.50150       4.00000      -1.00000   \n",
       " 1   32231.50000      -0.34107      -0.00195       7.00000      -0.96694   \n",
       " 2   41644.88095      -0.33455      -1.50247      12.00000      -1.00000   \n",
       " 3    -980.22431      -0.47359       7.48900       2.00000      -1.00000   \n",
       " 4   27792.27569      -0.18900      -4.50247       4.00000      -1.00000   \n",
       " \n",
       "    alpha_191_17  alpha_191_18  alpha_191_20  alpha_191_22  alpha_191_24  \\\n",
       " 0       0.06452       1.00098       0.04882      -0.00004       1.88100   \n",
       " 1       0.00826       1.00171       0.12210       0.00008       2.90480   \n",
       " 2       0.00027       1.00293       0.31762       0.00027       4.72384   \n",
       " 3       0.25402       1.00049       0.02443       0.00010       4.17907   \n",
       " 4       0.06452       1.00098       0.09773      -0.00003       4.14326   \n",
       " \n",
       "    alpha_191_25  alpha_191_26  alpha_191_28  alpha_191_29  alpha_191_31  \\\n",
       " 0           NaN           NaN    -885.70762       9.05687       0.14659   \n",
       " 1           NaN           NaN    -792.29149      32.97924       0.14248   \n",
       " 2           NaN           NaN    -677.01887      76.31029       0.25026   \n",
       " 3           NaN           NaN    -590.58457      11.20078      -0.02238   \n",
       " 4           NaN           NaN    -385.83657      35.70486       0.02034   \n",
       " \n",
       "    alpha_191_32  alpha_191_33  alpha_191_34  alpha_191_35  alpha_191_36  \\\n",
       " 0      -1.80488           NaN       0.99854      -0.55645       0.96639   \n",
       " 1      -2.16260           NaN       0.99858      -0.77419       0.97479   \n",
       " 2      -2.44715           NaN       0.99750      -0.76613       1.00000   \n",
       " 3      -2.11382           NaN       1.00022      -0.87097       0.96639   \n",
       " 4      -1.65854           NaN       0.99980      -0.45968       0.90756   \n",
       " \n",
       "    alpha_191_37  alpha_191_39  alpha_191_41  alpha_191_42  alpha_191_44  \\\n",
       " 0      -0.48182      -0.03220      -0.38934      -0.06120      14.00000   \n",
       " 1      -0.49091      -0.18667      -0.81967      -0.13718      15.00000   \n",
       " 2      -0.66364      -0.28830      -0.91803      -0.38616      19.00000   \n",
       " 3      -0.49091      -0.12976      -0.38115      -0.55910      15.00000   \n",
       " 4      -0.37273       0.05723      -0.38115      -0.46255      10.00000   \n",
       " \n",
       "    alpha_191_45  alpha_191_46  alpha_191_47  alpha_191_48  alpha_191_52  \\\n",
       " 0           NaN       0.99866      35.60726      -0.10288     132.38866   \n",
       " 1           NaN       0.99871      34.12003      -0.23808     138.93443   \n",
       " 2           NaN       0.99778      31.77817      -0.30610     141.90871   \n",
       " 3           NaN       1.00041      35.51221      -0.27579     121.01449   \n",
       " 4           NaN       0.99992      38.23308      -0.31933     102.95082   \n",
       " \n",
       "    alpha_191_54  alpha_191_57  alpha_191_60  alpha_191_61  alpha_191_62  \\\n",
       " 0      -0.85345      71.68610   38406.37619      -0.86239      -0.87745   \n",
       " 1      -0.51724      73.71666   31330.37619      -0.87156      -0.97633   \n",
       " 2      -0.87931      78.12994   37006.24286      -0.88991      -0.93843   \n",
       " 3      -0.03448      63.62509    7188.13759      -0.89908      -0.91333   \n",
       " 4      -0.60345      56.51929   55021.13759      -0.91743      -0.35267   \n",
       " \n",
       "    alpha_191_63  alpha_191_64  alpha_191_65  alpha_191_66  alpha_191_67  \\\n",
       " 0      71.31234      -0.78125       0.99890       0.10990      59.37147   \n",
       " 1      73.36048      -0.81967       0.99886       0.11395      59.78433   \n",
       " 2      82.40500      -0.95082       0.99793       0.20744      62.67685   \n",
       " 3      47.17123      -0.78689       1.00065      -0.06508      55.58965   \n",
       " 4      51.67898      -0.78125       1.00033      -0.03253      56.08239   \n",
       " \n",
       "    alpha_191_68  alpha_191_71  alpha_191_72  alpha_191_73  alpha_191_74  \\\n",
       " 0       0.00027       0.18229      36.04333      -4.13821       1.50765   \n",
       " 1       0.00037       0.19448      35.12313      -4.07317       1.47514   \n",
       " 2       0.00055       0.32579      33.64209      -4.04065       1.51626   \n",
       " 3       0.00036       0.05498      35.75657      -4.26829       1.49071   \n",
       " 4      -0.00001       0.10180      37.37324      -4.75610       1.52268   \n",
       " \n",
       "    alpha_191_76  alpha_191_77  alpha_191_78  alpha_191_79  alpha_191_80  \\\n",
       " 0       0.78738       0.10400      89.73294      64.54475      31.90899   \n",
       " 1       0.79173       0.11200     107.52928      65.60996     174.68728   \n",
       " 2       0.75093       0.10400     138.40304      71.26352     274.06196   \n",
       " 3       0.73413       0.41600      62.83142      53.63246     343.05314   \n",
       " 4       0.79176       0.18400     -26.41108      55.79945     321.00714   \n",
       " \n",
       "    alpha_191_81  alpha_191_82  alpha_191_83  alpha_191_85  alpha_191_87  \\\n",
       " 0   12858.67204      36.85575      -0.86777      50.00000      -7.50000   \n",
       " 1   14206.48986      36.13418      -0.99174     120.00000      -7.54545   \n",
       " 2   15141.52811      34.90384      -0.99174      19.00000      -7.73554   \n",
       " 3   18066.69779      36.41373      -0.96694     120.00000      -7.70248   \n",
       " 4   19825.57719      37.59679      -0.47107      85.50000      -5.62810   \n",
       " \n",
       "    alpha_191_88  alpha_191_89  alpha_191_90  alpha_191_91  alpha_191_92  \\\n",
       " 0       0.09768      -0.46202      -0.95041      -0.25815      -8.00000   \n",
       " 1       0.09766      -0.02909      -0.96694      -0.35649      -8.00000   \n",
       " 2       0.26862       0.92354      -1.00000      -0.41795      -8.00000   \n",
       " 3       0.07331      -0.01989      -0.86777      -0.34420      -8.00000   \n",
       " 4       0.31832      -0.44249      -0.36364      -0.16595      -8.00000   \n",
       " \n",
       "    alpha_191_95  alpha_191_96  alpha_191_97  alpha_191_99  alpha_191_100  \\\n",
       " 0    4363.14490      68.16023    3673.77679      -0.83471     4363.14490   \n",
       " 1    5072.65498      70.01237    6083.69270      -0.96694     5072.65498   \n",
       " 2    5593.83768      72.71823    6926.34825      -0.99174     5593.83768   \n",
       " 3    9042.13329      69.68718   11843.19057      -0.91736     9042.13329   \n",
       " 4   10163.31720      65.29789   13137.91015      -0.47934    10163.31720   \n",
       " \n",
       "    alpha_191_70  alpha_191_102  alpha_191_104  alpha_191_105  alpha_191_106  \\\n",
       " 0    4335.06867       64.74394        0.00428       -0.14967        4.00000   \n",
       " 1    7807.09945       74.61821       -0.01354       -0.39725        4.00000   \n",
       " 2    8593.64996       66.70759        0.00426       -0.55338       11.00000   \n",
       " 3   13582.15165       82.75614       -0.00608       -0.66473        3.00000   \n",
       " 4   13130.55269       66.36468        0.04960       -0.72153       13.00000   \n",
       " \n",
       "    alpha_191_107  alpha_191_108  alpha_191_109  alpha_191_110  alpha_191_111  \\\n",
       " 0       -0.31165       -0.93600        0.99192       79.48718    -4090.59717   \n",
       " 1       -0.31738       -0.96400        1.02253       82.05128     -329.65093   \n",
       " 2       -0.12239       -0.98800        1.06352       97.29730    -1309.32586   \n",
       " 3       -0.20535       -1.00000        1.29389       92.59259    12475.99145   \n",
       " 4       -0.00266       -0.81200        1.20825       97.40260    -6162.36601   \n",
       " \n",
       "    alpha_191_113  alpha_191_115  alpha_191_118  alpha_191_119  alpha_191_120  \\\n",
       " 0       -0.63852        0.72120       79.48718       -0.03725        0.51852   \n",
       " 1       -0.66851        0.84517       82.05128       -0.17156        0.51639   \n",
       " 2        0.69608        1.00000       97.29730       -0.11259        0.50400   \n",
       " 3        0.69118        1.00000       92.59259       -0.16344        0.58333   \n",
       " 4        0.68288        0.82004       97.40260       -0.26456        0.55022   \n",
       " \n",
       "    alpha_191_121  alpha_191_122  alpha_191_125  alpha_191_126  alpha_191_130  \\\n",
       " 0       -0.97200        0.00002        0.15045     4097.33333        0.96315   \n",
       " 1       -0.97600        0.00002        0.10832     4100.00000        0.93966   \n",
       " 2       -1.00000        0.00002        0.07822     4105.00000        0.93966   \n",
       " 3       -0.74650        0.00002        0.05004     4100.00000        0.98082   \n",
       " 4       -0.76858        0.00002        0.01404     4094.00000        0.99691   \n",
       " \n",
       "    alpha_191_131  alpha_191_134  alpha_191_135  alpha_191_136  alpha_191_138  \\\n",
       " 0        0.46135       45.37295        1.00307       -0.13894        5.52459   \n",
       " 1        0.17230       92.54528        1.00296       -0.37646        4.43443   \n",
       " 2        0.39270      111.69415        1.00287       -0.57170        2.19672   \n",
       " 3        0.00000       22.40704        1.00286       -0.02229        0.20492   \n",
       " 4        0.01633       26.77211        1.00274       -0.41517        0.36066   \n",
       " \n",
       "    alpha_191_139  alpha_191_140  alpha_191_141  alpha_191_142  alpha_191_145  \\\n",
       " 0       -0.18475        0.54400       -0.40777       -0.64223      -28.76837   \n",
       " 1       -0.61557        0.53600       -0.42718       -0.21306      -17.47967   \n",
       " 2       -0.72058        0.46400       -0.63107       -0.47663       -4.25877   \n",
       " 3       -0.89888        0.48800       -0.73786       -0.01687       12.18644   \n",
       " 4       -0.77285        0.19200       -0.59223       -0.37802       21.13633   \n",
       " \n",
       "    alpha_191_146   alpha_191_150  alpha_191_132  alpha_191_151  alpha_191_152  \\\n",
       " 0       -6.53767  76017825.33333    12974.70000       12.03946        0.00052   \n",
       " 1       -0.39126 110741000.00000    13263.80000       11.61458        0.00038   \n",
       " 2        2.80107  98626730.00000    13881.90000       11.60675        0.00024   \n",
       " 3        2.83704 188009600.00000    15217.10000       11.19016        0.00012   \n",
       " 4        1.53424 149574290.00000    15994.75000       11.24400        0.00002   \n",
       " \n",
       "    alpha_191_153  alpha_191_155  alpha_191_156  alpha_191_157  alpha_191_158  \\\n",
       " 0     4093.51042      277.32037       -0.48780        2.54622        0.00171   \n",
       " 1     4094.71875     1234.98624       -0.62602        4.42437        0.00195   \n",
       " 2     4096.89583     1588.46793       -0.83740        4.20168        0.00219   \n",
       " 3     4096.66667     3127.07653       -0.71951        2.19748        0.00464   \n",
       " 4     4096.66667     3332.05102       -0.83740        3.54622        0.00220   \n",
       " \n",
       "    alpha_191_159  alpha_191_161  alpha_191_162  alpha_191_163  alpha_191_168  \\\n",
       " 0   -60983.13769        6.58333        1.11289        0.29245       -1.42994   \n",
       " 1   -60463.01495        6.58333        1.12285        0.36792       -2.03637   \n",
       " 2   -58364.91617        6.91667        1.21680        0.09434       -1.73074   \n",
       " 3   -48680.07245        7.83333        0.86490        0.97170       -3.01345   \n",
       " 4   -47457.21300        8.16667        0.91720        0.50472       -2.28419   \n",
       " \n",
       "    alpha_191_169  alpha_191_170  alpha_191_173  alpha_191_175  alpha_191_176  \\\n",
       " 0       -0.02291       -0.57385     4107.99915        6.66667        0.80783   \n",
       " 1       -0.05585       -0.66761     4109.82784        6.83333        0.75798   \n",
       " 2       -0.07849       -0.86001     4113.58660        7.16667        0.78438   \n",
       " 3       -0.09006       -0.05791     4111.73888        9.66667        0.11551   \n",
       " 4       -0.09873       -0.57159     4111.04504        9.83333       -0.15960   \n",
       " \n",
       "    alpha_191_179  alpha_191_181  alpha_191_184  alpha_191_185  alpha_191_188  \\\n",
       " 0        0.42055       -0.06065            NaN        0.28000        3.40753   \n",
       " 1        0.54590       -0.07742            NaN        0.95200       14.39863   \n",
       " 2        0.64603       -0.09882            NaN        0.28800       22.31613   \n",
       " 3        0.27793       -0.11512            NaN        0.06400      100.53377   \n",
       " 4        0.08361       -0.11473            NaN        0.76800       -4.13704   \n",
       " \n",
       "    alpha_191_189  alpha_191_191  \n",
       " 0        1.63889       -2.45691  \n",
       " 1        2.08333       -0.23250  \n",
       " 2        3.33333       -1.24948  \n",
       " 3        3.75000        7.77465  \n",
       " 4        3.80556       -5.12675  ,\n",
       "                      date tic       open       high        low      close  \\\n",
       " 6428  2023-01-31 22:30:00  RB 4380.00000 4385.00000 4380.00000 4381.00000   \n",
       " 6429  2023-01-31 22:35:00  RB 4381.00000 4385.00000 4378.00000 4384.00000   \n",
       " 6430  2023-01-31 22:40:00  RB 4384.00000 4385.00000 4373.00000 4375.00000   \n",
       " 6431  2023-01-31 22:45:00  RB 4375.00000 4379.00000 4374.00000 4376.00000   \n",
       " 6432  2023-01-31 22:50:00  RB 4376.00000 4387.00000 4375.00000 4386.00000   \n",
       " \n",
       "           volume       open_oi      close_oi  alpha_191_1  alpha_191_2  \\\n",
       " 6428 13899.00000 1950293.00000 1952769.00000     -0.66485     -0.00000   \n",
       " 6429  9314.00000 1952769.00000 1953389.00000     -0.56531     -1.31429   \n",
       " 6430 13838.00000 1953389.00000 1952630.00000     -0.25831      1.38095   \n",
       " 6431  8422.00000 1952630.00000 1951328.00000      0.11301     -0.46667   \n",
       " 6432 14454.00000 1951328.00000 1950359.00000     -0.07761     -1.03333   \n",
       " \n",
       "       alpha_191_5  alpha_191_6  alpha_191_7  alpha_191_8  alpha_191_9  \\\n",
       " 6428     -0.66667     -0.83058      0.18177      0.05785      0.00072   \n",
       " 6429     -0.80178     -0.82645      0.08262      0.16529      0.00030   \n",
       " 6430     -0.80178     -0.82231      0.62793      0.60331     -0.00040   \n",
       " 6431     -0.80178     -0.32231      0.27266      0.53719     -0.00071   \n",
       " 6432     -0.62738     -0.32231      0.76013      0.28099      0.00056   \n",
       " \n",
       "       alpha_191_11  alpha_191_12  alpha_191_13  alpha_191_14  alpha_191_16  \\\n",
       " 6428   17342.59780      -0.48228       1.49929      10.00000      -0.97521   \n",
       " 6429   20577.45495      -0.45838      -2.50140      17.00000      -0.97521   \n",
       " 6430   12371.26447      -0.48662       3.99589      -1.00000      -0.97521   \n",
       " 6431   -1314.90476      -0.29979       0.49929      -6.00000      -0.95868   \n",
       " 6432   -7983.47619      -0.28024      -5.00411       6.00000      -0.90909   \n",
       " \n",
       "       alpha_191_17  alpha_191_18  alpha_191_20  alpha_191_22  alpha_191_24  \\\n",
       " 6428       0.00106       1.00229       0.25172       0.00041       6.04662   \n",
       " 6429       0.00001       1.00389       0.29741       0.00031       8.23729   \n",
       " 6430       1.98413       0.99977       0.18319       0.00009       6.38984   \n",
       " 6431      61.01221       0.99863       0.00000      -0.00008       3.91187   \n",
       " 6432       0.01639       1.00137       0.09128      -0.00007       4.32949   \n",
       " \n",
       "       alpha_191_25  alpha_191_26  alpha_191_28  alpha_191_29  alpha_191_31  \\\n",
       " 6428           NaN           NaN    -315.08298      34.98604       0.22114   \n",
       " 6429           NaN           NaN    -351.91538      27.70121       0.25919   \n",
       " 6430           NaN           NaN    -389.18363      25.35013       0.02477   \n",
       " 6431           NaN           NaN     -73.61578       0.00000       0.03048   \n",
       " 6432           NaN           NaN      62.53179      13.19398       0.21707   \n",
       " \n",
       "       alpha_191_32  alpha_191_33  alpha_191_34  alpha_191_35  alpha_191_36  \\\n",
       " 6428      -1.33884           NaN       0.99779      -0.91935       0.96639   \n",
       " 6429      -0.74380           NaN       0.99741      -0.91935       0.92437   \n",
       " 6430      -1.03306           NaN       0.99975      -0.91597       0.74790   \n",
       " 6431           NaN           NaN       0.99970      -0.80645       0.66387   \n",
       " 6432           NaN           NaN       0.99783      -0.82258       0.77311   \n",
       " \n",
       "       alpha_191_37  alpha_191_39  alpha_191_41  alpha_191_42  alpha_191_44  \\\n",
       " 6428      -0.87273      -0.40618      -0.39754      -0.37946      19.00000   \n",
       " 6429      -0.89091      -0.42651      -0.40164      -0.29142      17.00000   \n",
       " 6430      -0.60909      -0.32488      -0.40164      -0.21622      15.00000   \n",
       " 6431      -0.21818      -0.13789      -0.40164      -0.18632      14.00000   \n",
       " 6432      -0.56364      -0.30862      -0.40164      -0.16926      10.00000   \n",
       " \n",
       "       alpha_191_45  alpha_191_46  alpha_191_47  alpha_191_48  alpha_191_52  \\\n",
       " 6428           NaN       0.99884      37.09069      -0.13753      79.00000   \n",
       " 6429           NaN       0.99836      33.49860      -0.14722      71.11650   \n",
       " 6430           NaN       1.00042      35.62449      -0.14102      67.36111   \n",
       " 6431           NaN       1.00011      39.99954      -0.14637      69.90521   \n",
       " 6432           NaN       0.99800      36.34882      -0.16095      80.29557   \n",
       " \n",
       "       alpha_191_54  alpha_191_57  alpha_191_60  alpha_191_61  alpha_191_62  \\\n",
       " 6428      -0.62931      80.03636  -35376.45169      -0.92742      -0.57835   \n",
       " 6429      -0.75862      85.10360   -8692.86728      -0.95161       0.70565   \n",
       " 6430      -0.06897      74.19605    8047.79939      -0.76613       0.80039   \n",
       " 6431      -0.57759      68.51165   30317.11368      -0.76613      -0.67394   \n",
       " 6432      -0.97414      77.55849   24334.33590      -0.91935      -0.76520   \n",
       " \n",
       "       alpha_191_63  alpha_191_64  alpha_191_65  alpha_191_66  alpha_191_67  \\\n",
       " 6428      64.96866      -0.91803       0.99890       0.11045      45.73888   \n",
       " 6429      69.59312      -0.78125       0.99871       0.12943      47.39399   \n",
       " 6430      47.17431      -0.78125       1.00107      -0.10655      43.36198   \n",
       " 6431      49.34985      -0.78125       1.00084      -0.08372      43.47806   \n",
       " 6432      66.10213      -0.78125       0.99871       0.12937      48.78683   \n",
       " \n",
       "       alpha_191_68  alpha_191_71  alpha_191_72  alpha_191_73  alpha_191_74  \\\n",
       " 6428       0.00032       0.13428      44.81673      -4.20325       1.46653   \n",
       " 6429       0.00018       0.21526      42.14596      -4.42276       1.39126   \n",
       " 6430      -0.00014       0.02858      42.84641      -4.45528       1.06598   \n",
       " 6431      -0.00032       0.06384      44.98857      -4.37398       1.33210   \n",
       " 6432       0.00023       0.29537      42.47399      -4.75610       1.29822   \n",
       " \n",
       "       alpha_191_76  alpha_191_77  alpha_191_78  alpha_191_79  alpha_191_80  \\\n",
       " 6428       0.73820       0.28000     106.48829      53.16755     103.32066   \n",
       " 6429       0.73862       0.16800      93.55361      55.91501      30.55789   \n",
       " 6430       0.69390       0.34400      45.10476      46.90923     -37.91556   \n",
       " 6431       0.69096       0.39200      21.17081      47.92395     -67.85373   \n",
       " 6432       0.68371       0.18400      72.58922      56.92343      16.68685   \n",
       " \n",
       "       alpha_191_81  alpha_191_82  alpha_191_83  alpha_191_85  alpha_191_87  \\\n",
       " 6428   16984.50468      48.24383      -0.87603      44.00000      -4.45868   \n",
       " 6429   16253.92375      46.06601      -0.22314      24.00000      -4.95041   \n",
       " 6430   16023.83743      46.40608      -0.28926      98.00000      -4.88430   \n",
       " 6431   15299.85626      47.82398      -0.59504      72.00000      -4.80165   \n",
       " 6432   15219.27756      45.86067      -0.70248     108.00000      -4.79339   \n",
       " \n",
       "       alpha_191_88  alpha_191_89  alpha_191_90  alpha_191_91  alpha_191_92  \\\n",
       " 6428      -0.18227       6.31075      -0.90909      -0.50400      -8.00000   \n",
       " 6429       0.06848       6.70470      -0.38017      -0.47327      -8.00000   \n",
       " 6430      -0.06852       5.57236      -0.62810      -0.24585      -8.00000   \n",
       " 6431       0.22904       4.80612      -0.54545      -0.32576      -8.00000   \n",
       " 6432       0.27435       5.44690      -0.70248      -0.32576      -8.00000   \n",
       " \n",
       "       alpha_191_95  alpha_191_96  alpha_191_97  alpha_191_99  alpha_191_100  \\\n",
       " 6428   23077.10420      71.90518    6609.35430      -0.94215    23077.10420   \n",
       " 6429   23217.83696      76.30466    6573.65299      -0.46281    23217.83696   \n",
       " 6430   22849.35989      75.60179    6319.57835      -0.62810    22849.35989   \n",
       " 6431    7107.04625      73.23841    6496.31410      -0.53719     7107.04625   \n",
       " 6432    5664.42467      74.67844    6404.61453      -0.68595     5664.42467   \n",
       " \n",
       "       alpha_191_70  alpha_191_102  alpha_191_104  alpha_191_105  \\\n",
       " 6428    7935.35155       47.85387       -0.01181       -0.31855   \n",
       " 6429    7490.56424       42.48409        0.08681       -0.14447   \n",
       " 6430    6477.61817       49.22960        0.14910       -0.08268   \n",
       " 6431    6395.71599       42.13110        0.02649       -0.10686   \n",
       " 6432    2575.13407       51.48116       -0.00596       -0.08232   \n",
       " \n",
       "       alpha_191_106  alpha_191_107  alpha_191_108  alpha_191_109  \\\n",
       " 6428       -8.00000       -0.05480       -0.14400        0.86268   \n",
       " 6429        3.00000       -0.05535       -0.14800        0.88840   \n",
       " 6430       -3.00000       -0.34750       -0.15200        1.01402   \n",
       " 6431       10.00000       -0.02009       -0.11600        0.95029   \n",
       " 6432       12.00000       -0.12179       -0.17600        1.05509   \n",
       " \n",
       "       alpha_191_110  alpha_191_111  alpha_191_113  alpha_191_115  \\\n",
       " 6428       88.23529     3746.93465       -0.00206        0.25079   \n",
       " 6429       95.83333       25.31647        0.00701        0.24577   \n",
       " 6430       84.31373     3552.68632        0.00701        0.31747   \n",
       " 6431      106.25000     2273.34190        0.00211        0.34887   \n",
       " 6432      112.00000    -2825.13917       -0.00706        0.36403   \n",
       " \n",
       "       alpha_191_118  alpha_191_119  alpha_191_120  alpha_191_121  \\\n",
       " 6428       88.23529       -0.33560        3.40541       -0.00324   \n",
       " 6429       95.83333       -0.28475        3.00000       -0.16800   \n",
       " 6430       84.31373       -0.06986        4.20000       -0.12000   \n",
       " 6431      106.25000        0.00951        3.81818       -0.13200   \n",
       " 6432      112.00000        0.06014        2.62500       -0.19200   \n",
       " \n",
       "       alpha_191_122  alpha_191_125  alpha_191_126  alpha_191_130  \\\n",
       " 6428       -0.00002        0.00992     4382.00000        0.94547   \n",
       " 6429       -0.00002        0.00984     4382.33333        1.01503   \n",
       " 6430       -0.00002        0.01037     4377.66667        1.06973   \n",
       " 6431       -0.00001        0.01066     4376.33333        1.21782   \n",
       " 6432       -0.00001        0.01066     4382.66667        1.20879   \n",
       " \n",
       "       alpha_191_131  alpha_191_134  alpha_191_135  alpha_191_136  \\\n",
       " 6428        0.00017       60.54127        0.99418       -0.04915   \n",
       " 6429        0.01532       34.11722        0.99437       -0.03369   \n",
       " 6430        0.00000       47.60780        0.99469       -0.00571   \n",
       " 6431        0.00054       17.35700        0.99492        0.00757   \n",
       " 6432        0.70734       72.86618        0.99530        0.06636   \n",
       " \n",
       "       alpha_191_138  alpha_191_139  alpha_191_140  alpha_191_141  \\\n",
       " 6428        6.06557       -0.28321        0.45600       -0.04854   \n",
       " 6429        6.04918       -0.10190        0.38400       -0.01942   \n",
       " 6430        6.07377       -0.02765        0.43200       -0.03883   \n",
       " 6431        6.09016        0.01454        0.46400       -0.07767   \n",
       " 6432        6.11475        0.07872        0.28000       -0.12621   \n",
       " \n",
       "       alpha_191_142  alpha_191_145  alpha_191_146  alpha_191_150  \\\n",
       " 6428       -0.35539      -45.10931        1.11833 60905418.00000   \n",
       " 6429       -0.09431      -42.69273        3.66667 40817052.66667   \n",
       " 6430       -0.02305      -42.85546       -8.28341 60578151.33333   \n",
       " 6431       -0.35144      -44.49035        0.92214 36857479.33333   \n",
       " 6432       -0.76672      -37.61563       15.62283 63347064.00000   \n",
       " \n",
       "       alpha_191_132  alpha_191_151  alpha_191_152  alpha_191_153  \\\n",
       " 6428    21357.80000      -24.80672       -0.00029     4375.90625   \n",
       " 6429    20599.40000      -23.43471       -0.00028     4376.81250   \n",
       " 6430    19343.85000      -22.39923       -0.00025     4376.83333   \n",
       " 6431    14175.75000      -20.77927       -0.00019     4376.46875   \n",
       " 6432    13275.95000      -19.08533       -0.00010     4377.22917   \n",
       " \n",
       "       alpha_191_155  alpha_191_156  alpha_191_157  alpha_191_158  \\\n",
       " 6428      -29.81319       -0.92683        3.16807        0.00114   \n",
       " 6429     -308.88302       -0.95122        3.06723        0.00160   \n",
       " 6430     -159.93725       -0.82927        5.07143        0.00274   \n",
       " 6431     -384.83013       -0.65854        1.48319        0.00114   \n",
       " 6432     -101.00161       -0.69106        2.28571        0.00274   \n",
       " \n",
       "       alpha_191_159  alpha_191_161  alpha_191_162  alpha_191_163  \\\n",
       " 6428   -59905.27853        7.00000        1.42931        0.22642   \n",
       " 6429   -58036.73238        6.91667        1.46056        0.27358   \n",
       " 6430   -54485.64119        7.08333        1.14599        0.92453   \n",
       " 6431   -60182.58468        6.83333        1.16262        0.33019   \n",
       " 6432   -55194.19476        7.41667        1.40043        0.16038   \n",
       " \n",
       "       alpha_191_168  alpha_191_169  alpha_191_170  alpha_191_173  \\\n",
       " 6428       -0.65077       -0.04977       -0.57824     4383.99870   \n",
       " 6429       -0.45215        0.00935       -0.90362     4387.62570   \n",
       " 6430       -0.71537        0.08014       -0.03887     4386.60735   \n",
       " 6431       -0.59411        0.16198       -0.24206     4386.21615   \n",
       " 6432       -1.08874        0.24234       -0.67260     4390.03471   \n",
       " \n",
       "       alpha_191_175  alpha_191_176  alpha_191_179  alpha_191_181  \\\n",
       " 6428        6.83333        0.89784        0.20807        0.05445   \n",
       " 6429        7.33333        0.59926        0.25019        0.05700   \n",
       " 6430        8.16667        0.09274        0.15889        0.05708   \n",
       " 6431        6.83333        0.32202        0.35712        0.06475   \n",
       " 6432        7.66667        0.16712        0.55158        0.07237   \n",
       " \n",
       "       alpha_191_184  alpha_191_185  alpha_191_188  alpha_191_189  \\\n",
       " 6428            NaN        0.76000      -29.28165        5.22222   \n",
       " 6429            NaN        0.48800       -0.81500        5.50000   \n",
       " 6430            NaN        0.08800       50.82668        6.08333   \n",
       " 6431            NaN        0.75200      -32.60247        5.63889   \n",
       " 6432            NaN        0.08000       45.42566        5.02778   \n",
       " \n",
       "       alpha_191_191  \n",
       " 6428        2.46053  \n",
       " 6429       -1.74626  \n",
       " 6430        4.83853  \n",
       " 6431        1.21554  \n",
       " 6432       -4.38446  )"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "for i in range(columns_to_drop_1.shape[0]):\n",
    "    tmp_index = columns_to_drop_1[i]\n",
    "    tmp_column = columns_1[tmp_index]\n",
    "    res_df = res_df.drop(columns=tmp_column)\n",
    "\n",
    "res_df.isna().sum(), res_df.shape, res_df.columns, res_df.head(5), res_df.tail(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1ef4dc50",
   "metadata": {},
   "source": [
    "# Only Keep Valid Features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "aa1cd5d1",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-04-27T03:16:16.724747Z",
     "start_time": "2023-04-27T03:16:16.714742Z"
    },
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "def IC_test(price, factor_values, shift=-1):\n",
    "    next_term_return = pd.Series(price).pct_change().shift(shift).fillna(0).values\n",
    "    factor_values = np.asarray(factor_values)\n",
    "    factor_values = np.nan_to_num(factor_values)\n",
    "    IC = np.corrcoef(factor_values, next_term_return)[0][-1]\n",
    "    return IC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "72e49fc6",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-04-27T03:16:19.703695Z",
     "start_time": "2023-04-27T03:16:16.731751Z"
    },
    "code_folding": [
     4
    ],
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 129/129 [00:02<00:00, 44.06it/s]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "((123, 2),\n",
       "      correlation        factors\n",
       " 44       0.02678   alpha_191_65\n",
       " 77       0.02622  alpha_191_107\n",
       " 33       0.02371   alpha_191_46\n",
       " 24       0.02346   alpha_191_34\n",
       " 86       0.02052  alpha_191_120\n",
       " 72       0.02009   alpha_191_70\n",
       " 89       0.01999  alpha_191_125\n",
       " 108      0.01983  alpha_191_155\n",
       " 9        0.01944   alpha_191_13\n",
       " 29       0.01832   alpha_191_41\n",
       " 128      0.01753  alpha_191_191\n",
       " 87       0.01708  alpha_191_121\n",
       " 118      0.01698  alpha_191_170\n",
       " 115      0.01619  alpha_191_163\n",
       " 79       0.01519  alpha_191_109\n",
       " 65       0.01515   alpha_191_91\n",
       " 81       0.01487  alpha_191_111\n",
       " 37       0.01360   alpha_191_54\n",
       " 30       0.01305   alpha_191_42\n",
       " 78       0.01304  alpha_191_108\n",
       " 101      0.01286  alpha_191_145\n",
       " 121      0.01234  alpha_191_176\n",
       " 125      0.01229  alpha_191_185\n",
       " 56       0.01206   alpha_191_80\n",
       " 12       0.01161   alpha_191_17\n",
       " 100      0.01133  alpha_191_142\n",
       " 95       0.01063  alpha_191_136\n",
       " 61       0.01051   alpha_191_87\n",
       " 34       0.01042   alpha_191_47\n",
       " 84       0.01031  alpha_191_118\n",
       " 73       0.01030  alpha_191_102\n",
       " 27       0.01001   alpha_191_37\n",
       " 3        0.00989    alpha_191_6\n",
       " 74       0.00985  alpha_191_104\n",
       " 28       0.00959   alpha_191_39\n",
       " 5        0.00926    alpha_191_8\n",
       " 50       0.00916   alpha_191_73\n",
       " 4        0.00894    alpha_191_7\n",
       " 49       0.00814   alpha_191_72\n",
       " 41       0.00813   alpha_191_62\n",
       " 8        0.00738   alpha_191_12\n",
       " 58       0.00716   alpha_191_82\n",
       " 69       0.00703   alpha_191_97\n",
       " 1        0.00649    alpha_191_2\n",
       " 60       0.00617   alpha_191_85\n",
       " 22       0.00599   alpha_191_32\n",
       " 106      0.00594  alpha_191_152\n",
       " 123      0.00541  alpha_191_181\n",
       " 59       0.00533   alpha_191_83\n",
       " 64       0.00506   alpha_191_90,\n",
       " array([['alpha_191_65', 'alpha_191_107', 'alpha_191_46', 'alpha_191_34',\n",
       "         'alpha_191_120', 'alpha_191_70', 'alpha_191_125',\n",
       "         'alpha_191_155', 'alpha_191_13', 'alpha_191_41', 'alpha_191_191',\n",
       "         'alpha_191_121', 'alpha_191_170', 'alpha_191_163',\n",
       "         'alpha_191_109', 'alpha_191_91', 'alpha_191_111', 'alpha_191_54',\n",
       "         'alpha_191_42', 'alpha_191_108', 'alpha_191_145',\n",
       "         'alpha_191_176', 'alpha_191_185', 'alpha_191_80', 'alpha_191_17',\n",
       "         'alpha_191_142', 'alpha_191_136', 'alpha_191_87', 'alpha_191_47',\n",
       "         'alpha_191_118', 'alpha_191_102', 'alpha_191_37', 'alpha_191_6',\n",
       "         'alpha_191_104', 'alpha_191_39', 'alpha_191_8', 'alpha_191_73',\n",
       "         'alpha_191_7', 'alpha_191_72', 'alpha_191_62', 'alpha_191_12',\n",
       "         'alpha_191_82', 'alpha_191_97', 'alpha_191_2', 'alpha_191_85',\n",
       "         'alpha_191_32', 'alpha_191_152', 'alpha_191_181', 'alpha_191_83',\n",
       "         'alpha_191_90', 'alpha_191_110', 'alpha_191_188', 'alpha_191_35',\n",
       "         'alpha_191_36', 'alpha_191_189', 'alpha_191_1', 'alpha_191_150',\n",
       "         'alpha_191_169', 'alpha_191_100', 'alpha_191_95',\n",
       "         'alpha_191_146', 'alpha_191_138', 'alpha_191_139',\n",
       "         'alpha_191_74', 'alpha_191_105', 'alpha_191_135',\n",
       "         'alpha_191_157', 'alpha_191_158', 'alpha_191_119',\n",
       "         'alpha_191_151', 'alpha_191_122', 'alpha_191_156',\n",
       "         'alpha_191_175', 'alpha_191_159', 'alpha_191_28',\n",
       "         'alpha_191_113', 'alpha_191_115', 'alpha_191_130',\n",
       "         'alpha_191_52', 'alpha_191_141', 'alpha_191_81', 'alpha_191_99',\n",
       "         'alpha_191_140', 'alpha_191_16', 'alpha_191_78', 'alpha_191_88',\n",
       "         'alpha_191_106', 'alpha_191_131', 'alpha_191_153',\n",
       "         'alpha_191_61', 'alpha_191_173', 'alpha_191_96', 'alpha_191_11',\n",
       "         'alpha_191_126', 'alpha_191_60', 'alpha_191_132', 'alpha_191_77',\n",
       "         'alpha_191_57', 'alpha_191_76', 'alpha_191_44', 'alpha_191_79',\n",
       "         'alpha_191_67', 'alpha_191_48', 'alpha_191_168', 'alpha_191_63',\n",
       "         'alpha_191_24', 'alpha_191_161', 'alpha_191_5', 'alpha_191_64',\n",
       "         'alpha_191_162', 'alpha_191_89', 'alpha_191_71', 'alpha_191_68',\n",
       "         'alpha_191_179', 'alpha_191_20', 'alpha_191_22', 'alpha_191_14',\n",
       "         'alpha_191_18', 'alpha_191_9', 'alpha_191_134', 'alpha_191_31',\n",
       "         'alpha_191_29', 'alpha_191_66']], dtype=object))"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "factors = res_df.columns[9:]\n",
    "tickers = res_df.tic.unique()\n",
    "corr_list = []\n",
    "corr_df = pd.DataFrame()\n",
    "for i in tqdm(range(len(factors))):\n",
    "    tmp_corr_2 = []\n",
    "    for a in tickers:\n",
    "        tmp_df = res_df[res_df.tic==a]\n",
    "        tmp_corr = IC_test(price=tmp_df.close.values, factor_values=tmp_df[factors[i]].values, shift=-1)\n",
    "        tmp_corr_2.append(tmp_corr)\n",
    "    corr_list.append(np.mean(tmp_corr))\n",
    "        \n",
    "corr_df['correlation'] = corr_list\n",
    "corr_df['factors'] = factors\n",
    "corr_df = corr_df.dropna()\n",
    "corr_df = corr_df.drop(corr_df[corr_df.correlation==1].index)\n",
    "corr_df = corr_df.sort_values(by=\"correlation\",ascending=False)\n",
    "valid_factors = []\n",
    "valid_factors.append(corr_df.factors.values)\n",
    "valid_factors = np.asarray(valid_factors)\n",
    "\n",
    "corr_df.shape, corr_df.head(50), valid_factors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "c8ee7dc6",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-04-27T03:16:21.491899Z",
     "start_time": "2023-04-27T03:16:19.707698Z"
    }
   },
   "outputs": [],
   "source": [
    "corr_df.to_csv(path+'新建文件夹\\\\corr_df_alpha_191.csv')\n",
    "res_df.to_csv(path+'新建文件夹\\\\pre_processed_train_df_2_alpha_191.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "12b7e30d",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.13"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  },
  "vscode": {
   "interpreter": {
    "hash": "71159c3bea1635b2cdba277ecffbc628f93ffdcebf06bf6a363dd68ab8387877"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}