Price_Prediction.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "917-cSEM7PtP",
    "outputId": "8e5578ab-62b7-43cf-82f2-a5bf6073ed2e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
      "Requirement already satisfied: pandas in /usr/local/lib/python3.8/dist-packages (1.5.3)\n",
      "Requirement already satisfied: python-dateutil>=2.8.1 in /usr/local/lib/python3.8/dist-packages (from pandas) (2.8.2)\n",
      "Requirement already satisfied: numpy>=1.20.3 in /usr/local/lib/python3.8/dist-packages (from pandas) (1.22.4)\n",
      "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.8/dist-packages (from pandas) (2022.7.1)\n",
      "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.8/dist-packages (from python-dateutil>=2.8.1->pandas) (1.15.0)\n",
      "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
      "Requirement already satisfied: pandas-datareader in /usr/local/lib/python3.8/dist-packages (0.10.0)\n",
      "Requirement already satisfied: pandas>=0.23 in /usr/local/lib/python3.8/dist-packages (from pandas-datareader) (1.5.3)\n",
      "Requirement already satisfied: requests>=2.19.0 in /usr/local/lib/python3.8/dist-packages (from pandas-datareader) (2.28.2)\n",
      "Requirement already satisfied: lxml in /usr/local/lib/python3.8/dist-packages (from pandas-datareader) (4.9.2)\n",
      "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.8/dist-packages (from pandas>=0.23->pandas-datareader) (2022.7.1)\n",
      "Requirement already satisfied: numpy>=1.20.3 in /usr/local/lib/python3.8/dist-packages (from pandas>=0.23->pandas-datareader) (1.22.4)\n",
      "Requirement already satisfied: python-dateutil>=2.8.1 in /usr/local/lib/python3.8/dist-packages (from pandas>=0.23->pandas-datareader) (2.8.2)\n",
      "Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.8/dist-packages (from requests>=2.19.0->pandas-datareader) (3.0.1)\n",
      "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.8/dist-packages (from requests>=2.19.0->pandas-datareader) (2.10)\n",
      "Requirement already satisfied: urllib3<1.27,>=1.21.1 in /usr/local/lib/python3.8/dist-packages (from requests>=2.19.0->pandas-datareader) (1.26.14)\n",
      "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.8/dist-packages (from requests>=2.19.0->pandas-datareader) (2022.12.7)\n",
      "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.8/dist-packages (from python-dateutil>=2.8.1->pandas>=0.23->pandas-datareader) (1.15.0)\n",
      "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
      "Requirement already satisfied: yfinance in /usr/local/lib/python3.8/dist-packages (0.2.12)\n",
      "Requirement already satisfied: multitasking>=0.0.7 in /usr/local/lib/python3.8/dist-packages (from yfinance) (0.0.11)\n",
      "Requirement already satisfied: appdirs>=1.4.4 in /usr/local/lib/python3.8/dist-packages (from yfinance) (1.4.4)\n",
      "Requirement already satisfied: requests>=2.26 in /usr/local/lib/python3.8/dist-packages (from yfinance) (2.28.2)\n",
      "Requirement already satisfied: html5lib>=1.1 in /usr/local/lib/python3.8/dist-packages (from yfinance) (1.1)\n",
      "Requirement already satisfied: lxml>=4.9.1 in /usr/local/lib/python3.8/dist-packages (from yfinance) (4.9.2)\n",
      "Requirement already satisfied: cryptography>=3.3.2 in /usr/local/lib/python3.8/dist-packages (from yfinance) (39.0.1)\n",
      "Requirement already satisfied: pandas>=1.3.0 in /usr/local/lib/python3.8/dist-packages (from yfinance) (1.5.3)\n",
      "Requirement already satisfied: beautifulsoup4>=4.11.1 in /usr/local/lib/python3.8/dist-packages (from yfinance) (4.11.2)\n",
      "Requirement already satisfied: numpy>=1.16.5 in /usr/local/lib/python3.8/dist-packages (from yfinance) (1.22.4)\n",
      "Requirement already satisfied: pytz>=2022.5 in /usr/local/lib/python3.8/dist-packages (from yfinance) (2022.7.1)\n",
      "Requirement already satisfied: frozendict>=2.3.4 in /usr/local/lib/python3.8/dist-packages (from yfinance) (2.3.5)\n",
      "Requirement already satisfied: soupsieve>1.2 in /usr/local/lib/python3.8/dist-packages (from beautifulsoup4>=4.11.1->yfinance) (2.4)\n",
      "Requirement already satisfied: cffi>=1.12 in /usr/local/lib/python3.8/dist-packages (from cryptography>=3.3.2->yfinance) (1.15.1)\n",
      "Requirement already satisfied: six>=1.9 in /usr/local/lib/python3.8/dist-packages (from html5lib>=1.1->yfinance) (1.15.0)\n",
      "Requirement already satisfied: webencodings in /usr/local/lib/python3.8/dist-packages (from html5lib>=1.1->yfinance) (0.5.1)\n",
      "Requirement already satisfied: python-dateutil>=2.8.1 in /usr/local/lib/python3.8/dist-packages (from pandas>=1.3.0->yfinance) (2.8.2)\n",
      "Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.8/dist-packages (from requests>=2.26->yfinance) (3.0.1)\n",
      "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.8/dist-packages (from requests>=2.26->yfinance) (2.10)\n",
      "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.8/dist-packages (from requests>=2.26->yfinance) (2022.12.7)\n",
      "Requirement already satisfied: urllib3<1.27,>=1.21.1 in /usr/local/lib/python3.8/dist-packages (from requests>=2.26->yfinance) (1.26.14)\n",
      "Requirement already satisfied: pycparser in /usr/local/lib/python3.8/dist-packages (from cffi>=1.12->cryptography>=3.3.2->yfinance) (2.21)\n",
      "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
      "Requirement already satisfied: stockstats in /usr/local/lib/python3.8/dist-packages (0.5.2)\n",
      "Requirement already satisfied: pandas>=0.24.2 in /usr/local/lib/python3.8/dist-packages (from stockstats) (1.5.3)\n",
      "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.8/dist-packages (from pandas>=0.24.2->stockstats) (2022.7.1)\n",
      "Requirement already satisfied: python-dateutil>=2.8.1 in /usr/local/lib/python3.8/dist-packages (from pandas>=0.24.2->stockstats) (2.8.2)\n",
      "Requirement already satisfied: numpy>=1.20.3 in /usr/local/lib/python3.8/dist-packages (from pandas>=0.24.2->stockstats) (1.22.4)\n",
      "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.8/dist-packages (from python-dateutil>=2.8.1->pandas>=0.24.2->stockstats) (1.15.0)\n",
      "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
      "Requirement already satisfied: matplotlib in /usr/local/lib/python3.8/dist-packages (3.5.3)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.8/dist-packages (from matplotlib) (1.4.4)\n",
      "Requirement already satisfied: numpy>=1.17 in /usr/local/lib/python3.8/dist-packages (from matplotlib) (1.22.4)\n",
      "Requirement already satisfied: pillow>=6.2.0 in /usr/local/lib/python3.8/dist-packages (from matplotlib) (8.4.0)\n",
      "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.8/dist-packages (from matplotlib) (0.11.0)\n",
      "Requirement already satisfied: python-dateutil>=2.7 in /usr/local/lib/python3.8/dist-packages (from matplotlib) (2.8.2)\n",
      "Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.8/dist-packages (from matplotlib) (4.38.0)\n",
      "Requirement already satisfied: pyparsing>=2.2.1 in /usr/local/lib/python3.8/dist-packages (from matplotlib) (3.0.9)\n",
      "Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.8/dist-packages (from matplotlib) (23.0)\n",
      "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.8/dist-packages (from python-dateutil>=2.7->matplotlib) (1.15.0)\n",
      "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
      "Requirement already satisfied: PyPortfolioOpt in /usr/local/lib/python3.8/dist-packages (1.5.4)\n",
      "Requirement already satisfied: pandas>=0.19 in /usr/local/lib/python3.8/dist-packages (from PyPortfolioOpt) (1.5.3)\n",
      "Requirement already satisfied: numpy<2.0.0,>=1.22.4 in /usr/local/lib/python3.8/dist-packages (from PyPortfolioOpt) (1.22.4)\n",
      "Requirement already satisfied: scipy<2.0,>=1.3 in /usr/local/lib/python3.8/dist-packages (from PyPortfolioOpt) (1.7.3)\n",
      "Requirement already satisfied: cvxpy<2.0.0,>=1.1.10 in /usr/local/lib/python3.8/dist-packages (from PyPortfolioOpt) (1.2.3)\n",
      "Requirement already satisfied: osqp>=0.4.1 in /usr/local/lib/python3.8/dist-packages (from cvxpy<2.0.0,>=1.1.10->PyPortfolioOpt) (0.6.2.post0)\n",
      "Requirement already satisfied: setuptools<=64.0.2 in /usr/local/lib/python3.8/dist-packages (from cvxpy<2.0.0,>=1.1.10->PyPortfolioOpt) (57.4.0)\n",
      "Requirement already satisfied: scs>=1.1.6 in /usr/local/lib/python3.8/dist-packages (from cvxpy<2.0.0,>=1.1.10->PyPortfolioOpt) (3.2.2)\n",
      "Requirement already satisfied: ecos>=2 in /usr/local/lib/python3.8/dist-packages (from cvxpy<2.0.0,>=1.1.10->PyPortfolioOpt) (2.0.12)\n",
      "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.8/dist-packages (from pandas>=0.19->PyPortfolioOpt) (2022.7.1)\n",
      "Requirement already satisfied: python-dateutil>=2.8.1 in /usr/local/lib/python3.8/dist-packages (from pandas>=0.19->PyPortfolioOpt) (2.8.2)\n",
      "Requirement already satisfied: qdldl in /usr/local/lib/python3.8/dist-packages (from osqp>=0.4.1->cvxpy<2.0.0,>=1.1.10->PyPortfolioOpt) (0.1.5.post3)\n",
      "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.8/dist-packages (from python-dateutil>=2.8.1->pandas>=0.19->PyPortfolioOpt) (1.15.0)\n"
     ]
    }
   ],
   "source": [
    "!pip install --upgrade pandas\n",
    "!pip install --upgrade pandas-datareader\n",
    "!pip install yfinance\n",
    "!pip install stockstats\n",
    "!pip install matplotlib\n",
    "!pip install PyPortfolioOpt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "ohZgJtt9Af0_"
   },
   "outputs": [],
   "source": [
    "# pandas, pandas_datareader, numpy and matplotlib\n",
    "\n",
    "import numpy as np\n",
    "from matplotlib import pyplot\n",
    "import pandas as pd\n",
    "from pandas.plotting import scatter_matrix\n",
    "import seaborn as sns\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from pandas.plotting import scatter_matrix\n",
    "from statsmodels.graphics.tsaplots import plot_acf\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "yrSmcs3t7yas"
   },
   "outputs": [],
   "source": [
    "from stockstats import StockDataFrame\n",
    "import pandas_datareader.data as pdr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 741
    },
    "id": "RMnZE3cD26J1",
    "outputId": "21443938-0c42-4446-dd4c-30481bc3c4fe"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[*********************100%***********************]  7 of 7 completed\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-6ee07240-4aaa-4ee9-9f2e-4b6b05c43271\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"7\" halign=\"left\">Adj Close</th>\n",
       "      <th colspan=\"3\" halign=\"left\">Close</th>\n",
       "      <th>...</th>\n",
       "      <th colspan=\"3\" halign=\"left\">Open</th>\n",
       "      <th colspan=\"7\" halign=\"left\">Volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>AAPL</th>\n",
       "      <th>AMZN</th>\n",
       "      <th>CSCO</th>\n",
       "      <th>GOOGL</th>\n",
       "      <th>HD</th>\n",
       "      <th>IBM</th>\n",
       "      <th>MSFT</th>\n",
       "      <th>AAPL</th>\n",
       "      <th>AMZN</th>\n",
       "      <th>CSCO</th>\n",
       "      <th>...</th>\n",
       "      <th>HD</th>\n",
       "      <th>IBM</th>\n",
       "      <th>MSFT</th>\n",
       "      <th>AAPL</th>\n",
       "      <th>AMZN</th>\n",
       "      <th>CSCO</th>\n",
       "      <th>GOOGL</th>\n",
       "      <th>HD</th>\n",
       "      <th>IBM</th>\n",
       "      <th>MSFT</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1962-01-02</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.613219</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.374124</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>407940</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1962-01-03</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.627321</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.291268</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>305955</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1962-01-04</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.611102</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.355003</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>274575</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1962-01-05</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.579375</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.272148</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>384405</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1962-01-08</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.549760</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.131931</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>572685</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-02-22</th>\n",
       "      <td>148.910004</td>\n",
       "      <td>95.790001</td>\n",
       "      <td>49.310001</td>\n",
       "      <td>91.650002</td>\n",
       "      <td>296.299988</td>\n",
       "      <td>130.970001</td>\n",
       "      <td>251.509995</td>\n",
       "      <td>148.910004</td>\n",
       "      <td>95.790001</td>\n",
       "      <td>49.310001</td>\n",
       "      <td>...</td>\n",
       "      <td>297.549988</td>\n",
       "      <td>131.899994</td>\n",
       "      <td>254.089996</td>\n",
       "      <td>51011300.0</td>\n",
       "      <td>59534100.0</td>\n",
       "      <td>20301900.0</td>\n",
       "      <td>30884000.0</td>\n",
       "      <td>5922100.0</td>\n",
       "      <td>3200200</td>\n",
       "      <td>22491100.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-02-23</th>\n",
       "      <td>149.399994</td>\n",
       "      <td>95.820000</td>\n",
       "      <td>49.209999</td>\n",
       "      <td>90.889999</td>\n",
       "      <td>299.309998</td>\n",
       "      <td>130.789993</td>\n",
       "      <td>254.770004</td>\n",
       "      <td>149.399994</td>\n",
       "      <td>95.820000</td>\n",
       "      <td>49.209999</td>\n",
       "      <td>...</td>\n",
       "      <td>296.790009</td>\n",
       "      <td>131.500000</td>\n",
       "      <td>255.559998</td>\n",
       "      <td>48394200.0</td>\n",
       "      <td>48467000.0</td>\n",
       "      <td>14066200.0</td>\n",
       "      <td>41206400.0</td>\n",
       "      <td>4354500.0</td>\n",
       "      <td>3725600</td>\n",
       "      <td>29219100.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-02-24</th>\n",
       "      <td>146.710007</td>\n",
       "      <td>93.500000</td>\n",
       "      <td>48.480000</td>\n",
       "      <td>89.129997</td>\n",
       "      <td>296.660004</td>\n",
       "      <td>130.570007</td>\n",
       "      <td>249.220001</td>\n",
       "      <td>146.710007</td>\n",
       "      <td>93.500000</td>\n",
       "      <td>48.480000</td>\n",
       "      <td>...</td>\n",
       "      <td>292.399994</td>\n",
       "      <td>129.619995</td>\n",
       "      <td>249.960007</td>\n",
       "      <td>55418200.0</td>\n",
       "      <td>57010100.0</td>\n",
       "      <td>17251300.0</td>\n",
       "      <td>36554600.0</td>\n",
       "      <td>3713000.0</td>\n",
       "      <td>3015800</td>\n",
       "      <td>24978700.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-02-27</th>\n",
       "      <td>147.919998</td>\n",
       "      <td>93.760002</td>\n",
       "      <td>48.730000</td>\n",
       "      <td>89.870003</td>\n",
       "      <td>296.010010</td>\n",
       "      <td>130.490005</td>\n",
       "      <td>250.160004</td>\n",
       "      <td>147.919998</td>\n",
       "      <td>93.760002</td>\n",
       "      <td>48.730000</td>\n",
       "      <td>...</td>\n",
       "      <td>299.359985</td>\n",
       "      <td>131.419998</td>\n",
       "      <td>252.460007</td>\n",
       "      <td>44998500.0</td>\n",
       "      <td>47470300.0</td>\n",
       "      <td>14776400.0</td>\n",
       "      <td>27502300.0</td>\n",
       "      <td>3238300.0</td>\n",
       "      <td>2761300</td>\n",
       "      <td>21190000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-02-28</th>\n",
       "      <td>147.410004</td>\n",
       "      <td>94.230003</td>\n",
       "      <td>48.419998</td>\n",
       "      <td>90.059998</td>\n",
       "      <td>296.540009</td>\n",
       "      <td>129.300003</td>\n",
       "      <td>249.419998</td>\n",
       "      <td>147.410004</td>\n",
       "      <td>94.230003</td>\n",
       "      <td>48.419998</td>\n",
       "      <td>...</td>\n",
       "      <td>295.000000</td>\n",
       "      <td>130.550003</td>\n",
       "      <td>249.070007</td>\n",
       "      <td>50455400.0</td>\n",
       "      <td>43901300.0</td>\n",
       "      <td>16621200.0</td>\n",
       "      <td>30081100.0</td>\n",
       "      <td>3948600.0</td>\n",
       "      <td>5141900</td>\n",
       "      <td>22467600.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>15395 rows × 42 columns</p>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-6ee07240-4aaa-4ee9-9f2e-4b6b05c43271')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-6ee07240-4aaa-4ee9-9f2e-4b6b05c43271 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-6ee07240-4aaa-4ee9-9f2e-4b6b05c43271');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "             Adj Close                                               \\\n",
       "                  AAPL       AMZN       CSCO      GOOGL          HD   \n",
       "Date                                                                  \n",
       "1962-01-02         NaN        NaN        NaN        NaN         NaN   \n",
       "1962-01-03         NaN        NaN        NaN        NaN         NaN   \n",
       "1962-01-04         NaN        NaN        NaN        NaN         NaN   \n",
       "1962-01-05         NaN        NaN        NaN        NaN         NaN   \n",
       "1962-01-08         NaN        NaN        NaN        NaN         NaN   \n",
       "...                ...        ...        ...        ...         ...   \n",
       "2023-02-22  148.910004  95.790001  49.310001  91.650002  296.299988   \n",
       "2023-02-23  149.399994  95.820000  49.209999  90.889999  299.309998   \n",
       "2023-02-24  146.710007  93.500000  48.480000  89.129997  296.660004   \n",
       "2023-02-27  147.919998  93.760002  48.730000  89.870003  296.010010   \n",
       "2023-02-28  147.410004  94.230003  48.419998  90.059998  296.540009   \n",
       "\n",
       "                                         Close                        ...  \\\n",
       "                   IBM        MSFT        AAPL       AMZN       CSCO  ...   \n",
       "Date                                                                  ...   \n",
       "1962-01-02    1.613219         NaN         NaN        NaN        NaN  ...   \n",
       "1962-01-03    1.627321         NaN         NaN        NaN        NaN  ...   \n",
       "1962-01-04    1.611102         NaN         NaN        NaN        NaN  ...   \n",
       "1962-01-05    1.579375         NaN         NaN        NaN        NaN  ...   \n",
       "1962-01-08    1.549760         NaN         NaN        NaN        NaN  ...   \n",
       "...                ...         ...         ...        ...        ...  ...   \n",
       "2023-02-22  130.970001  251.509995  148.910004  95.790001  49.310001  ...   \n",
       "2023-02-23  130.789993  254.770004  149.399994  95.820000  49.209999  ...   \n",
       "2023-02-24  130.570007  249.220001  146.710007  93.500000  48.480000  ...   \n",
       "2023-02-27  130.490005  250.160004  147.919998  93.760002  48.730000  ...   \n",
       "2023-02-28  129.300003  249.419998  147.410004  94.230003  48.419998  ...   \n",
       "\n",
       "                  Open                              Volume              \\\n",
       "                    HD         IBM        MSFT        AAPL        AMZN   \n",
       "Date                                                                     \n",
       "1962-01-02         NaN    7.374124         NaN         NaN         NaN   \n",
       "1962-01-03         NaN    7.291268         NaN         NaN         NaN   \n",
       "1962-01-04         NaN    7.355003         NaN         NaN         NaN   \n",
       "1962-01-05         NaN    7.272148         NaN         NaN         NaN   \n",
       "1962-01-08         NaN    7.131931         NaN         NaN         NaN   \n",
       "...                ...         ...         ...         ...         ...   \n",
       "2023-02-22  297.549988  131.899994  254.089996  51011300.0  59534100.0   \n",
       "2023-02-23  296.790009  131.500000  255.559998  48394200.0  48467000.0   \n",
       "2023-02-24  292.399994  129.619995  249.960007  55418200.0  57010100.0   \n",
       "2023-02-27  299.359985  131.419998  252.460007  44998500.0  47470300.0   \n",
       "2023-02-28  295.000000  130.550003  249.070007  50455400.0  43901300.0   \n",
       "\n",
       "                                                                    \n",
       "                  CSCO       GOOGL         HD      IBM        MSFT  \n",
       "Date                                                                \n",
       "1962-01-02         NaN         NaN        NaN   407940         NaN  \n",
       "1962-01-03         NaN         NaN        NaN   305955         NaN  \n",
       "1962-01-04         NaN         NaN        NaN   274575         NaN  \n",
       "1962-01-05         NaN         NaN        NaN   384405         NaN  \n",
       "1962-01-08         NaN         NaN        NaN   572685         NaN  \n",
       "...                ...         ...        ...      ...         ...  \n",
       "2023-02-22  20301900.0  30884000.0  5922100.0  3200200  22491100.0  \n",
       "2023-02-23  14066200.0  41206400.0  4354500.0  3725600  29219100.0  \n",
       "2023-02-24  17251300.0  36554600.0  3713000.0  3015800  24978700.0  \n",
       "2023-02-27  14776400.0  27502300.0  3238300.0  2761300  21190000.0  \n",
       "2023-02-28  16621200.0  30081100.0  3948600.0  5141900  22467600.0  \n",
       "\n",
       "[15395 rows x 42 columns]"
      ]
     },
     "execution_count": 295,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import yfinance as yfin\n",
    "yfin.pdr_override()\n",
    "stk_tickers = [\"MSFT\", \"IBM\", \"GOOGL\", \"AMZN\",\"AAPL\",\"CSCO\",\"HD\"]\n",
    "stk_data = pdr.get_data_yahoo(stk_tickers)\n",
    "stk_data\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ctaNY-y9UZnI"
   },
   "source": [
    "# Single stock example"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "bOhZUmJNU36e"
   },
   "source": [
    "## Params"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "VtzQB2807CwO"
   },
   "outputs": [],
   "source": [
    "# params change this to modify the whole thing\n",
    "pred_stock = 'MSFT'\n",
    "num_folds = 2\n",
    "scoring_mse = \"neg_mean_squared_error\"\n",
    "scoring_mae = \"mean_absolute_error\"\n",
    "scoring_r2_score = \"neg_r2_score\"\n",
    "scoring_ape = \"neg_mean_absolute_percentage_error\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "hI2qfAeILbNA"
   },
   "outputs": [],
   "source": [
    "from google.colab import files\n",
    "\n",
    "#stk_data.to_csv(\"stk_data.csv\")\n",
    "#files.download(\"stk_data.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "5z9R8YCKJ8fX",
    "outputId": "bdef49da-2486-4eff-db03-8a34bcc2f2c2"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['IBM', 'GOOGL', 'AMZN', 'AAPL', 'CSCO', 'HD']"
      ]
     },
     "execution_count": 298,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "interested_stock = stk_tickers.copy()\n",
    "interested_stock.remove(pred_stock)\n",
    "interested_stock"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "U4JXYO-U6YR2"
   },
   "outputs": [],
   "source": [
    "temp = pd.DataFrame({\"close\": stk_data.loc[:, ('Adj Close', pred_stock)],\n",
    "\"open\": stk_data.loc[:, ('Open', pred_stock)],\n",
    "\"high\" : stk_data.loc[:, ('High', pred_stock)],\n",
    "\"low\" : stk_data.loc[:, ('Low', pred_stock)],\n",
    "\"volume\" : stk_data.loc[:, ('Volume', pred_stock)]})\n",
    "pred_stockDataFrame = StockDataFrame(temp.copy())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "shLtvtK0AeB0"
   },
   "outputs": [],
   "source": [
    "for stock in interested_stock:\n",
    "  temp[stock] = pd.DataFrame({stock: stk_data.loc[:, ('Adj Close', stock)]})\n",
    "  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "6_PrmnKwolur"
   },
   "outputs": [],
   "source": [
    "temp['next_close'] = temp['close'].shift(-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "YOb6jSQoCJEN"
   },
   "outputs": [],
   "source": [
    "# Add a feature\n",
    "temp[\"movement percent\"] = np.where((temp[\"close\"]-temp[\"open\"])==0 , 0,  (temp[\"high\"]-temp[\"low\"])/(temp[\"close\"]-temp[\"open\"]))\n",
    "temp.drop([\"open\",\"high\",\"low\"],axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "57DfwUQABEeJ"
   },
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.ensemble import GradientBoostingRegressor\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "from sklearn.metrics import r2_score\n",
    "from sklearn.metrics import mean_absolute_percentage_error\n",
    "from sklearn.feature_selection import SelectKBest\n",
    "from sklearn.feature_selection import chi2, f_regression\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.linear_model import Ridge\n",
    "from sklearn.linear_model import Lasso\n",
    "from sklearn.linear_model import ElasticNet\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.neighbors import KNeighborsRegressor\n",
    "from sklearn.svm import SVR\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.ensemble import ExtraTreesRegressor\n",
    "from sklearn.ensemble import AdaBoostRegressor\n",
    "from sklearn.neural_network import MLPRegressor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "P7ECymS7UeD0"
   },
   "source": [
    "## models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "s2TP0GRopJde"
   },
   "outputs": [],
   "source": [
    "models = []\n",
    "# Do modelName, modelFunc, errorValueTemp\n",
    "models.append(('LR', LinearRegression(),np.inf))\n",
    "# Boosting methods\n",
    "models.append(('GBR', GradientBoostingRegressor(),np.inf))\n",
    "\n",
    "# Normally ridge would be one model but since it's a model that run fastest\n",
    "models.append(('R1',Ridge(alpha=1),np.inf))\n",
    "models.append(('R0.5',Ridge(alpha=0.5),np.inf))\n",
    "models.append(('R0.2',Ridge(alpha=0.2),np.inf))\n",
    "models.append(('R3',Ridge(alpha=3),np.inf))\n",
    "models.append(('LASSO', Lasso(),np.inf))\n",
    "models.append(('EN',ElasticNet(),np.inf))\n",
    "models.append(('KNN',KNeighborsRegressor(),np.inf))\n",
    "models.append(('CART',DecisionTreeRegressor(),np.inf))\n",
    "models.append(('SVR',SVR(),np.inf))\n",
    "# Boosting methods\n",
    "models.append(('ABR',AdaBoostRegressor(),np.inf))\n",
    "models.append(('GBR',GradientBoostingRegressor(),np.inf))\n",
    "# Bagging methods\n",
    "models.append(('RFR',RandomForestRegressor(),np.inf))\n",
    "models.append(('ETR',ExtraTreesRegressor(),np.inf))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "EuL0_9kspPZw"
   },
   "outputs": [],
   "source": [
    "used_data = temp.dropna().copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "OwCC-1v-GnDk"
   },
   "outputs": [],
   "source": [
    "# First separation, one for training and choosing, the other for testing\n",
    "used_data_xtrain, used_data_xtest, used_data_ytrain, used_data_ytest = train_test_split(used_data.drop(\"next_close\",axis=1), used_data[\"next_close\"], test_size=0.3, shuffle=False)\n",
    "trainAndEval = pd.concat([used_data_xtrain,used_data_ytrain],axis=1)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "JRms_cpgKBep",
    "outputId": "e3674dc5-098c-4f52-8e1f-08a5f58365c0"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LR: 0.123537 (0.036311) 0.107398 0.451148\n",
      "GBR: 0.688673 (0.244619) 0.068325 335.117410\n",
      "R1: 0.121118 (0.031198) 0.107406 0.450773\n",
      "R0.5: 0.121974 (0.033449) 0.107400 0.450931\n",
      "R0.2: 0.122804 (0.035071) 0.107398 0.451054\n",
      "R3: 0.121574 (0.025180) 0.107468 0.450710\n",
      "LASSO: 1.995804 (1.534931) 1.108537 79.696465\n",
      "EN: 2.557644 (1.376308) 1.366963 90.539966\n",
      "KNN: 3.043993 (0.186667) 0.140036 354.977294\n",
      "CART: 0.790324 (0.228495) 0.000000 312.898227\n",
      "SVR: 2.977605 (2.127604) 0.142390 539.242190\n",
      "ABR: 0.709375 (0.332396) 0.169906 349.151039\n",
      "GBR: 0.662035 (0.257486) 0.068325 337.411053\n",
      "RFR: 0.412476 (0.257786) 0.017011 328.444310\n",
      "ETR: 1.468660 (0.415519) 0.000000 328.438398\n"
     ]
    }
   ],
   "source": [
    "\n",
    "data = trainAndEval\n",
    "# Transform in case we use lasso\n",
    "scaler = StandardScaler()\n",
    "X_train, X_test, y_train, y_test = train_test_split(data.drop(\"next_close\",axis=1), data[\"next_close\"], test_size=0.4, shuffle=False)\n",
    "X_train_fit = pd.DataFrame(scaler.fit_transform(X_train),columns = X_train.columns)\n",
    "X_test_fit = pd.DataFrame(  scaler.transform(X_test),columns = X_test.columns)\n",
    "\n",
    "\n",
    "names = []\n",
    "kfold_results = []\n",
    "test_results = []\n",
    "train_results = []\n",
    "\n",
    "for name, model,err in models:\n",
    "    names.append(name)\n",
    "    ## k-fold analysis:\n",
    "    # kfold = KFold(n_splits=num_folds, random_state=seed)\n",
    "    kfold = KFold(n_splits=num_folds)\n",
    "    #converted mean squared error to positive. The lower the better\n",
    "    cv_results = -1* cross_val_score(model, X_train_fit, y_train, cv=kfold, scoring=scoring_mse)\n",
    "    kfold_results.append(cv_results)\n",
    "    # Full Training period\n",
    "    model.fit(X_train_fit, y_train)\n",
    "    train_result = mean_squared_error(model.predict(X_train_fit), y_train)\n",
    "    train_results.append(train_result)\n",
    "    # Test results\n",
    "    test_result = mean_squared_error(model.predict(X_test_fit), y_test)\n",
    "    test_results.append(test_result)\n",
    "    msg = \"%s: %f (%f) %f %f\" % (name, cv_results.mean(), cv_results.std(), train_result, test_result)\n",
    "    print(msg)\n",
    "res = pd.DataFrame({\"name\" :[x[0] for x in models], \"test_results\":test_results})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "id": "bQ6bDb7Th-cy",
    "outputId": "a81948da-4b41-4566-cccd-b95231cdf6ed"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "type": "string"
      },
      "text/plain": [
       "'R3'"
      ]
     },
     "execution_count": 308,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get the minimum mean squared error model\n",
    "minModelName = list(res[res[\"test_results\"]==min(res[\"test_results\"])][\"name\"])[0]\n",
    "minModelName"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "CFy5PuHWk6cU"
   },
   "outputs": [],
   "source": [
    "# Prediction on real test\n",
    "predictionsmse = []\n",
    "for name, model,err in models:\n",
    "  if name == minModelName:\n",
    "    predictionsmse = model.predict(pd.DataFrame(scaler.transform(used_data_xtest),columns = used_data_xtest.columns))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 540
    },
    "id": "e8_Tp-Fas8QL",
    "outputId": "29a14837-b411-4f87-d918-c581a4194d2a"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2AAAAILCAYAAACO1v+JAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/NK7nSAAAACXBIWXMAAAsTAAALEwEAmpwYAAApO0lEQVR4nO3deZRsWVkn7N8LxShQ3JSSwSooGwHFAdT72YrMKiDgtLSFalqBTxtxaETFme665YwtaqvYfIAIggiojQsFBRQRaNHmFhYlBaiMVjFIQV0skEEo3u+Pc5KOm2TezJs3c2dm5POsFSsjzrh3nMiI84u9z47q7gAAALD7rrXXBQAAADgsBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDCANarqaVX107u07YdU1YtPMf+eVXXFbuz7oKuqH6+qp+x1OU6lqr6rqv65qj5YVZ++ybIvq6rv2GDe+VXVVXXW7pR0a05VRgC2RwADDq355PJEVV1v1D67+3e6+z4LZeiq+uxR+6/Jo6rqdVX1r1V1RVX9XlV9wagybFd3/2x372kYWHu8quoxVfWuqvq8qrpOkl9Kcp/uvlF3v2/vSrrzquphVfXKvS4HwEEngAGHUlWdn+RuSTrJ1w3a5562Zsz+R5LvS/KoJCtJbp/kD5M8YA/LtKl98tydpKoem+TRSe7R3ZcluXmS6ye5bHA59t1zA8DGBDDgsPq2JH+d5GlJHnqqBavqh+dWjndW1XcstoJU1dlV9dtVdWVVvb2qHltV15rnPayq/ndV/XJVvS/JscVWhKp6+byL185d1h60sM8frKr3zPt9+ML0p1XVb1TVn8zr/O+qukVV/crcmvfGqvqiDepxuyTfk+SC7n5pd3+0uz80t8r9/GnW5/1V9Zaquss8/fK5vA9dU9YnVtVLquoDVfWXVXWbhfn/Y17v6qq6uKrutjDvWFX9flU9s6quTvKwedoz5/nXn+e9by7Lq6vq5vO8W1XV86vqqqp6U1X95zXbfe5cxw9U1WVVdfRUx3+D5/Knk3xHkrt39z9U1e2T/P08+/1V9dJ5ubvMZfuX+e9dNtjetavqF6vqvVX1lmwSiKvqbVX1I1V1aZJ/raqzqurLquqv5ufjtVV1z4XlHzYfrw9U1Vur6iELz8czF5Zbt+tjVX1ukicm+fL5dff+efr9q+r183bfUVWPOY2nEeBQEsCAw+rbkvzOfLvv6sn7WlV1vyQ/kOSrknx2knuuWeTXkpyd5N8luce83YcvzP/3Sd6SqXXkZxZX7O67z3fvNHdZe878+BbzNj8zybcneUJVHVlY9VuSPDbJzZJ8NMmrkrxmfvz7mbrBrecrk1zR3f9ng/lbrc+lST49ybOSPDvJ/5PpuflPSX69qm60sPxDkvzUXLZLMj3fq16d5M6ZWuKeleT3qur6C/O/fq7PTdesl0yh+ewk581leWSSD8/znp3kiiS3SvLNSX62qu69sO7XzcvcNMnzk/z66ow53P7Gek/Mgp9P8qBM4estSdLd/5Dk8+b5N+3ue1fVSpIXJPnVuYy/lOQFtf61Yf85yQOTfFGSo3O5N3NBpqB200yvrxck+elMz+djkvxBVZ1TVZ82l+FruvvGSe6S6VhsWXe/IdNz/Kr5tXrTedZvJvnOebufn+Slp7NdgMNIAAMOnaq6a5LbJHlud1+c5M1J/uMGi39Lkt/q7su6+0NJji1s59pJHpzkx7r7A939tiSPT/KtC+u/s7t/rbs/3t0fztZ8LMlPdvfHuvuFST6Y5A4L85/X3Rd390eSPC/JR7r7t7v7miTPyXQSv55PT/KujXa6xfq8tbt/a2Ff581l/Wh3vzjJv2UKY6te0N0v7+6PJvmJTC0o5yVJdz+zu983PzePT3K9NfV8VXf/YXd/Yp3n7mNzfT67u6+Zn4+r521/RZIf6e6PdPclSZ6SKUiuemV3v3CuwzOS3Gl1Rnd/d3d/90bP0ew+Sf60u/9pk+UekOQfu/sZcx1/N8kbk3ztOst+S5Jf6e7Lu/uqJD+3ybaT5Ffn5T+cKfy+cK7XJ7r7JUmOJ7n/vOwnknx+Vd2gu981d5ncCR9Lcsequkl3n+ju1+zQdgGWlgAGHEYPTfLi7n7v/PhZ2bgb4q2SXL7wePH+zZJcJ8nbF6a9PVPL1XrLb9X7uvvjC48/lGSxVemfF+5/eJ3Hi8uetN0ktzzFfrdSn7X7Snefav+frH93fzDJVZme09UBLN4wd897f6YWrZutt+46npHkRUmeXVPX0F+oaRCMWyW5qrs/cIo6vHvh/oeSXH9tl7tNPDjJN1fVRZssd6uc/FyuV5bFZS9fs9xmFpe/TZL/MHc/fP/8fN41yS27+18ztdg9Msm7quoFVfU5W9j+VnxTppD39rmL6Zfv0HYBlpYABhwqVXWDTK0N96iqd1fVu5N8f5I7VdWd1lnlXUnOXXh83sL992ZqAbjNwrRbJ3nHwuPekYLvjD9Pcu4prnnaSn1O1yefr7lr4kqSd87Xe/1wpmNxZO7S9i9JamHdDZ+7uXXwou6+Y6YudQ/M1Mr1ziQrVXXjHazDWv+QqUvqd1fVj55iuXfm5OfyVGV5V05+bd16C+VYfH4uT/KM7r7pwu3TVq/t6+4XdfdXZwrgb0zy5Hm9f01yw4Xt3GKL+8u83Vd399cn+YxMg7k8dwvlBjjUBDDgsPmGJNckuWOm64/unORzk7wiJ3dTW/XcJA+vqs+tqhsm+a+rM+YubM9N8jNVdeN5gIkfSPLMdbazkX/OdL3Vruvuf0zyG0l+t6bfG7vuPJjFg6vqR3eoPmvdv6ruWlXXzXQt2F939+VJbpzk40muTHJWVf23JDfZ6kar6l5V9QVzt8mrMwXHT8zb/qskPzfX7QszXUd3JnX4FHMXvq9K8kNV9egNFnthkttX1X+cB8l4UKbX3R+vs+xzkzyqqs6dr/c7VbBbzzOTfG1V3Xce0OP68zE+t6puXlVfP18L9tFMXVo/Ma93SZK7V9Wtq+rsJD92in38c6YAf90kmV8/D6mqs7v7Y5mOwydOsT4AEcCAw+ehma7p+qfufvfqLdNADA9Z2xWtu/8k0wAGf5HkTZlGTkymE9kk+S+ZWhHekuSVmbozPvU0ynMsydPnbmPfss06nY5HZarrE5K8P9P1b9+Y5I/m+Wdan7WeleTCTF0PvyTTtUrJ1H3wTzO1Jr09yUdyet01b5FpgI6rk7whyV9m6paYTINTnJ+pBep5SS7s7j/bykZrGrXxiVtZtrtfm+S+SS6sqkeuM/99mVrmfjBT988fTvLAha6vi56c6Tl5baYBVf7XVsqwsK/LMw1a8uOZQu3lSX4o0+f8tTIF6XdmOg73SPJd83ovyXQt36VJLs764XDVSzMNsf/uqlqtw7cmeVtNI1U+MtOgKwCcQnXvp94xAPvbPBz365Jcb811WqxRVU/LNOriY/e6LACwX2gBA9hEVX1jVV1v7hr2uCR/JHwBANshgAFs7juTvCdTd71rMnffAgA4XbogAgAADKIFDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYBABDAAAYJCzdmOjN7vZzfr888/fjU0DAADsexdffPF7u/uctdN3JYCdf/75OX78+G5sGgAAYN+rqrevN10XRAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEHO2usCAAAsm6ra1nrdvcMlAfYbAQwAYIdtFKSqSsiCQ04XRAAAgEEEMAAAgEEEMAAAgEG2dA1YVb0tyQeSXJPk4919dDcLBQAAsIxOZxCOe3X3e3etJAAAAEtOF0QAAIBBthrAOsmLq+riqnrEegtU1SOq6nhVHb/yyit3roQAAABLYqsB7K7d/cVJvibJ91TV3dcu0N1P6u6j3X30nHPO2dFCAgAALIMtBbDufsf89z1JnpfkS3ezUAAAAMto0wBWVZ9WVTdevZ/kPklet9sFAwAAWDZbGQXx5kmeV1Wryz+ru/90V0sFAACwhDYNYN39liR3GlAWAACApWYYegAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMACAbVhZWUlVndYtyWmvU1VZWVnZ49oCO+WsrS5YVddOcjzJO7r7gbtXJACA/e/EiRPp7iH7Wg1vwMF3Oi1g35fkDbtVEAAAgGW3pQBWVecmeUCSp+xucQAAAJbXVlvAfiXJDyf5xEYLVNUjqup4VR2/8sord6JsAAAAS2XTAFZVD0zynu6++FTLdfeTuvtodx8955xzdqyAAAAAy2IrLWBfkeTrquptSZ6d5N5V9cxdLRUAAMAS2jSAdfePdfe53X1+kgcneWl3/6ddLxkAAMCS8TtgAAAAg2z5d8CSpLtfluRlu1ISADgNZ/K7SKN+uwkA1jqtAAYA+8WpQlRVCVkA7Eu6IAIAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAwigAEAAAxy1l4XAIDdUVXbXre7d7AkAMAqAQxgSZ0qRFWVkAUAe0AXRAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAD2rZWVlVTVad+SbGu9lZWVPa4xAMvurL0uAABs5MSJE+nuYftbDW/7wZmUZeRzBsDpEcAAYB86VYiqKiEL4IDSBREAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwA9ogfmgY4fPwOGADskcP8Q9MAh5UWMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEE2HQWxqq6f5OVJrjcv//vdfeFuFwwAll1feJPk2Nlj9wfAntrKMPQfTXLv7v5gVV0nySur6k+6+693uWwAsNTqoquHD0Pfx4btDoB1bBrAevpk+OD88DrzbdynBQAAwJLY0jVgVXXtqrokyXuSvKS7/2ZXSwUAALCEthTAuvua7r5zknOTfGlVff7aZarqEVV1vKqOX3nllTtcTAAAgINvK9eAfVJ3v7+q/iLJ/ZK8bs28JyV5UpIcPXpUF0UAYKmNHETFACqwPLYyCuI5ST42h68bJPnqJI/b9ZIBAOxjIwdRMYAKLI+ttIDdMsnTq+rambosPre7/3h3iwUAh0NVDdvXkSNHhu0LgPVtZRTES5N80YCyAMChst3Wk6oaOnw9ADtnS4NwAAAAcOYEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEG28jtgAMBgm/0+2KnmG6IeYP8SwABgHxKiAJaTLogAAACDCGAAAACDCGAAAACDCGAAAACDCGAAAACDCGAAB9jKykqq6rRvSba13srKyh7XGAAONsPQAxxgJ06cGDpc+Wa/TQUAnJoWMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEEEMAAAgEHO2usCAMBG+sKbJMfOHrs/ANhFAhgA+1ZddHW6e9z+qtLHhu0OgENIF0QAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBNg1gVXVeVf1FVb2+qi6rqu8bUTAAAIBlc9YWlvl4kh/s7tdU1Y2TXFxVL+nu1+9y2QAAAJbKpi1g3f2u7n7NfP8DSd6Q5DN3u2AAAADL5rSuAauq85N8UZK/WWfeI6rqeFUdv/LKK3eoeAAAAMtjK10QkyRVdaMkf5Dk0d199dr53f2kJE9KkqNHj/aOlRCADfWFN0mOnT12fwDAtm0pgFXVdTKFr9/p7v+1u0UCYKvqoqvTPe47r6pKHxu2OwBYOlsZBbGS/GaSN3T3L+1+kQAAAJbTVq4B+4ok35rk3lV1yXy7/y6XCwAAYOls2gWxu1+ZpAaUBQAAYKmd1iiIAAAAbN+WR0EEAOBk06Xyu+/IkSND9gPsPgEMAGAbtjMCaVUNHbkU2H8EMAAAYCmcSav0qC9HBDAAAGApnCpE7ZcWaAEMgH1t1DU2ietsANh9AhgA+9Z2v6ncL99yAsBahqEHAAAYRAADAAAYRAADAAAYRAADAAAYRAADAAAYRAADAAAYRAADAAAYRAADAAAYRAADAAAYRAADAAAYRAADAAAYRAADAAAYRAADAAAYRAADAAAOjJWVlVTVad+SbGu9lZWVHS3/WTu6NQAAgF104sSJdPew/a2Gt52iBQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGCQs/a6AAAAAFvVF94kOXb22P3tIAEMAAA4MOqiq9Pd4/ZXlT62c9vTBREAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGCQs/a6AAAAAKejqobt68iRIzu6PQEMAAA4MLp7W+tV1bbX3UkCGMABd5C/BQSAw0YAAzjADvq3gABw2BiEAwAAYBABDAAAYBBdEAE4kDa79u1U83W/BGCvCGAAHEhCFAAHkS6IAAAAgwhgAAAAgwhgAAAAgwhgAAAAgwhgAAAAgwhgAAAAgwhgAAAAgwhgAAAAg/ghZgAAYClU1bbnd/dOF2ddmwawqnpqkgcmeU93f/7uFwkAgP1qsxPcUxl1gsvhdRBeY1vpgvi0JPfb5XIAAHAAdPeGt63Mh8Nu0xaw7n55VZ0/oCwAwCGhFQU4rHbsGrCqekSSRyTJrW99653aLACwhE4VoqpKyAKW1o6NgtjdT+ruo9199JxzztmpzQIAACwNw9ADAAAMIoABAAAMsmkAq6rfTfKqJHeoqiuq6tt3v1gAAADLZyujIF4woiCHmZGgAADgcNixURDZPiNBAQDsD74YZ7cJYAAAMPPFOLvNIBwAwK5YWVlJVZ32Lcm21ltZWdnjGgNsTgsYALArTpw4MbS14Ey6jgGMogUMAICTaL2E3aMFDACAk2i9hN0jgAEA7LBTBYpTzTPAAyw/AQwAYIcJUsBGXAMGAAAwiAA2iItZAQAAXRAHcTErAAAggAEAu6IvvEly7Oyx+4MtWFlZyYkTJ7a17na+5D5y5Eiuuuqqbe2P5SOAAQC7oi66enjvjz42bHccYHomsZdcAwYAADCIAAYAADCIAAYAADCIAAYAADCIQTgAADiJESxh9whgAACcxAiWsHsEsEF8kwQAAAhgg/gmCYDDaOTvHx05cmTYvuCgOpP/yZHnsstMAAMAdsV2T9aqyokeu+ow90w61f+W/70xBDAAAA4VPZPYS4ahBwAAGEQAAwAAGEQAAwAAGEQAAwAAGEQAAwAAGMQoiAP5LRQAADjcBLBB/BYKAPxfm30pear5PheBg0wAAwCGE6L2Pz13Dq6VlZWcOHFiW+tu57gfOXIkV1111bb2dxgJYAAAnOQw9NxZ5oB54sSJ4T80zdYJYAAAHCqHIWCyfxkFEQAAYBABDAAAYBABDAAAYBABDAAAYBCDcAAsKb+zBAD7jwC2DzhJAnaD9wcA2H8EsH3ASRIAABwOAhgAACyRvvAmybGzx+6PLRPAAABgtgyXhtRFVw8tS1Wljw3b3YEngAEAwGy/hCiWlwAGAMCWLUMLEewlAQwAgC0TouDM+CFmAACAQQQwAACAQQQwAACAQQQwAACAQQQwAACAQQQwAACAQQxDDwAAS2az32vbSUeOHBm2r2UggAEAwBLZ7m+1VZXfeRtAF0QAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBBDAAAIBBDszvgG33x+T8lgGcmTP5IUf/fwAAJzswAWyjEzk/GLf/LfsJ/LLX71RlXIb/P1/uAHCYbPa5d6r5Pvt2xr7qgriyspKqOq1bktNep6qysrKyx7VdLqc6dmdivxy/w1y/Zfj/243jt1/qBgCno7u3fWNn7KsWsKsedU2Smwza2zWD9nM4jD12yejjp347bZnr570FANhY7UaaPXr0aB8/fvz0CzOwO9MydJ3aT0Y/n/Znf6e7v1GOHDmSq666atj+AID9qaou7u6ja6fvqxawZNyJ0pEjR4bs5zAZfZI7mvrtnNH12+w6tp3eJgDARvZVAHOSdHDtxrHbbLsjqd/Ob3e/OAhlBACWx74KYKfiJOngWvZjp34AAGzVvhoFEQAAYJkJYAAAAINsKYBV1f2q6u+r6k1V9aO7XSgAAIBltGkAq6prJ3lCkq9JcsckF1TVHXe7YAAAAMtmKy1gX5rkTd39lu7+tyTPTvL1u1ssAACA5bOVAPaZSS5feHzFPO0kVfWIqjpeVcevvPLKnSofAADA0tixQTi6+0ndfbS7j55zzjk7tVkAAIClsZUA9o4k5y08PneeBgAAwGnYSgB7dZLbVdVnVdV1kzw4yfN3t1gAAADL56zNFujuj1fV9yZ5UZJrJ3lqd1+26yUDAABYMpsGsCTp7hcmeeEulwUAAGCp7dggHAAAAJyaAAYAADCIAAYAADCIAAYAADCIAAYAADCIAAYAADCIAAYAADBIdffOb7TqyiRv3/ENr+9mSd47aF97Qf0ONvU7uJa5bon6HXTqd3Atc90S9Tvo1G9n3aa7z1k7cVcC2EhVdby7j+51OXaL+h1s6ndwLXPdEvU76NTv4FrmuiXqd9Cp3xi6IAIAAAwigAEAAAyyDAHsSXtdgF2mfgeb+h1cy1y3RP0OOvU7uJa5bon6HXTqN8CBvwYMAADgoFiGFjAAAIADQQADAAAY5EAFsKr64DrTjlXVO6rqkqp6fVVdsBdl246qunlVPauq3lJVF1fVq6rqG6vqnlX1L3OdLq2qP6uqz5jXeVhVXTnPe2NVff9e12Mrquqaucyvq6o/qqqbLsz706p6f1X98R4W8bRsUp+HVtU/zreHbrD+4uv2kqq6/7DCb2IH6vbf59fmpVX1vMX196uN6lxVt6mq18zzLquqR+5xUU+y3nviwrxfmV9j11qYdvOq+uOqeu38fvnCefq1qupX5/r/XVW9uqo+a553dlX9dlW9qarePN8/e/drtzULx2719qPz9JdV1fGF5Y5W1cv2rKBbtHhMq+r+VfUP8+vwWFV9aPWzYJ1lu6oev/D4MVV1bFjBN1FVt6iqZ8+voYur6oVVdft53qOr6iOLr6s1n4NvrKpfrKovWDjOV1XVW+f7f7Z3NftUVfUT8/vFpXP5Lqyqn1uzzJ2r6g3z/bfN/3eXVtVfVtVt9qbkW1NV3zC/3j5nfnx+VX14rutrq+qvquoO87xPOY57W/pTqyU/LzvFZ93iMVy9Xfcg1S1Z//OgpvOQS+bPsH9ZmHeX+XPi7+fX7aur6s7DCtvdB+aW5IPrTDuW5DHz/dsluTrJdfa6rFuoSyV5VZJHLky7TZL/kuSeSf54YfrPJblovv+wJL8+3//0TD8md95e1+d0jl2Spyf5iYXHX5nkaxfrvN9vG9UnyUqSt8x/j8z3j6yz/idft/vttgN1u0+Ss+b7j0vyuL2u0xnU+bpJrjffv1GStyW51V6Xd71yr5l+rSRvT/LXSe61MP3/S/J9C4+/cP57QZLfT3Kt+fG5q8d2nn5sYZ2LkvzeXtd9C8/By5L8U5KvmR8fTfKyvS7vVuszvy++Kclt58fH5vo8bu2y8/2PJHlrkpvNjx+zeNz2uE7rfd7dKcnd5vt/k+QVSR6+MP+eq58JSW6Q5I1JvmJh/tOSfPNe122dun75XNfV942bJbl7kresWe7nk/y3+f7bFo7bRUmevNf12KSOz5mP1+p5yflJXrcw/zuTPH0rx3E/3TZ4nS7VeVk2/qw76RguLHNg6ra2fuvMO+kYztNeluTofP/hSV4yqqwHqgVsM939j0k+lOnkcL+7d5J/6+4nrk7o7rd3968tLlRVleTGSU6s3UB3vy/TB/Qtd7msO+1VST5z9UF3/3mSD+xdcc7YYn3um+kf+KruPpHkJUnut2clO3OnXbfufnF3f3x++NeZTuYPkk/Wubv/rbs/Ok+/Xg5Or4F7Jrksyf/MFK5W3TLJFasPuvvShenv6u5PzNOv6O4TVfXZSb4kyU8tbOMnkxytqtvuXvF3zH9P8hN7XYjTVVV3T/LkJA/s7jcvzHpqkgdV1co6q3080+he+/Eb6nsl+diaz7vXdvcr5tfRjZI8Nie/VrOw7IeTXJKFz4197JZJ3rv6vtHd7+3ulyc5UVX/fmG5b0nyu+usf9Ln435TVTdKctck357kwRssdpOsf86y34/jYTsvO63X2gGr23YM/d87KCcTW1JVX5zkH7v7PXtdli34vCSvOcX8u1XVJZm+8fyqTB+8J6mqWye5fpJL187br6rq2pm+2X3+XpdlJ6xTn89McvnCIldk43/o7527Mjy1qvbdlwZnWLdV/2+SP9n50u2O9V6fVXVeVV2aqe6P6+537lX5TsMFmU7unpfkAVV1nXn6E5L8ZlX9xdxN6lbz9Ocm+dq5W8bjq+qL5ul3THJJd1+zuuH5/iWZ3sP2gxus6XLyoIV5r0ryb1V1r70q3DZcL8kfJvmG7n7jmnkfzPRZ8H0brPuEJA+pfdRFdPb5SS7eYN6Dkzw7U4vKHarq5msXmN8fb5fk5btWwp3z4iTn1dR19Deq6h7z9N/NHFiq6suSXDV/abzW/TId//3q65P8aXf/Q5L3VdWXzNNvO///vTnJDyT5pbUrHoDjeGjOyzY4F7vtwvvoE9ZZ5yDU7VSfB5sZ+r+3LAHs+6vqskzdGH5mrwuzHVX1hNU+qPOkV3T3nbv7vCS/leQXFhZ/0HxC+KYkv9HdHxld3m24wfzG9e4kN8/UenKQnWl9/meS2ya5c5J3JXn8KZcea0eOVVX9RKZv5X9n54q2azasc3df3t1fmOSzkzx0vRPE/aSqrpvk/kn+sLuvzvS+eN8k6e4XJfl3mVpXPifJ31bVOd19RZI7JPmxJJ9I8udV9ZV7Uf5t+PD8Xrl6e86a+T+dqXXloPhYkr/K1MKwnl/N9Dq88doZ8/H+7SSP2r3i7bgLkjx7bn39gyT/YWHe3arqtUnekeRF3f3uvSjg6ejuD2ZqNX5EkiuTPKeqHpap294313RN5oPzqa1ff1FV70jyNevM208uyBSYM/9dbbV88/z/d9skj87Jv7V04I5jsrTnZaf6fH/zwvvo9yxMPyh1Szb/PFjP71TVWzP1lviU4LlbliWA/XJ3f16Sb8r07e7197pAW3BZki9efTC/2L8yyTnrLPv8TH3IVz1nPiG8S5Kfr6pb7GZBd8iHu/vOmfpTV5LvOfXi+95G9XlHkvMWljt3nnaS7v7n7r5mPul4cpIv3d3inpYzqlsyXZSc5IFJHtJz5+p9btPX59zy9bokdxtbtNN23yQ3TfJ3VfW2TN2FPtm1a+5C+qzu/tYkr8783tLdH+3uP+nuH0rys0m+Icnrk9y5Th7I41qZvjh4/YjKnKnufmmma0++bK/LskWfyNQ97Uur6sfXzuzu9yd5VjZ+D/2VTOHt03apfNtxWaZQcpKq+oJMLSIvmV+rD87J3RBf0d13ytQy8e1DL5A/A/N7+8u6+8Ik35vkm7r78kzX6N0j07nK2hPDe2V6/7kk03Vg+87c9fXeSZ4yH68fyvRarTWLrj1nOSjH8TCcl23nXOyg1G27HpLpi8mnJ/m1TZbdMcsSwJIk3f38JMeTrDs62z7z0iTXr6rvWph2ww2WvWuSN6+d2N3HkzwjG3dH2Xe6+0OZvp39wao6a6/Lc6bWqc+Lktynqo7M3S3uM087SVUt9qH+xkwn9vvKGdTtfkl+OMnXzds4MNbWuarOraobJJ/sPnPXJH+/l2XcgguSfEd3n9/d5yf5rCRfXVU3rKp7V9UNk2RuQbltkn+qqi9e7Y44B6wvTPL27n5Tkr/NyS1Ij03ymnneQfHTmV6TB8L8OnxApu6E67WE/VKmgQ4+5T20u6/K1KV0oxa0vfDSJNerqkesTqiqL8zUmnds9bXa3bdKcqtaMwpgd78106AVPzKy0NtRVXeoqtstTLpzpgFxkqll65czDchxxdp152tnH53k2za4zm+vfXOSZ3T3bebjdV6mUHnemuU2OmfZ78fx0JyXbedc7KDUbTvmL4r/a5Ivq3l0z9120ALYDavqioXbD6yzzE8m+YHFb2z3o/lgf0OSe9Q0lO7/yZS+V9+Y7jb3X31tkm9N8oMbbOpxSR6+XneU/aq7/zZTH+ILkqSqXpHk95J85Xxc77uX5Ttdi/WZT35+KlPLwquT/OQ8LVX1lKo6Oq/2CzUPO5zpm8/9eOH8duv265kuUH7J/Bp+4jqb3rfWvD4/N8nfzP+Hf5nkF7v77/ayfGusfU/88Uz92F+wukB3/2uSV2YaafRLkhyfX3evSvKU7n51ks9I8kdV9bpMdf94puOYTCfyt69p+PA3J7l99tfJ/do+/z+/doHufmGm7mAHxvy/db8kj62qr1sz772Zru+73garPz7T6Hv7wvx5941Jvmp+HV2WaRS5e2aqx6LnZf3BHZ6Y5O5Vdf4uFnUn3CjJ02v6mYdLM11HeWye93uZWoE27GLY3e+a5+/HXiIX5FOP1x9k6rq8ev3QazO1oH/HBtvYt8fxsJ2XrT0X26L9XrdNPw82Mg8S8/hMLbu7rg5G7yAAAICDb1+3EgEAACwTAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGAQAQwAAGCQ/x99Jr5NBi1xFwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = pyplot.figure()\n",
    "fig.suptitle('Algorithm Comparison: Kfold results')\n",
    "ax = fig.add_subplot(111)\n",
    "pyplot.boxplot(kfold_results)\n",
    "ax.set_xticklabels(names)\n",
    "fig.set_size_inches(15,8)\n",
    "pyplot.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "dicJIo-ClF69"
   },
   "source": [
    "MAE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "4kkI54iHhaq8",
    "outputId": "b7213124-e115-4cc0-aa64-0e1332f68e64"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LR: 1.468660 (0.415519) 0.223465 0.480420\n",
      "GBR: 1.468660 (0.415519) 0.190240 14.056156\n",
      "R1: 1.468660 (0.415519) 0.223569 0.479923\n",
      "R0.5: 1.468660 (0.415519) 0.223515 0.480151\n",
      "R0.2: 1.468660 (0.415519) 0.223484 0.480304\n",
      "R3: 1.468660 (0.415519) 0.223857 0.479580\n",
      "LASSO: 1.468660 (0.415519) 0.805424 7.575020\n",
      "EN: 1.468660 (0.415519) 0.848435 7.843710\n",
      "KNN: 1.468660 (0.415519) 0.250746 14.817387\n",
      "CART: 1.468660 (0.415519) 0.000000 13.762146\n",
      "SVR: 1.468660 (0.415519) 0.237153 19.294248\n",
      "ABR: 1.468660 (0.415519) 0.328074 14.592934\n",
      "GBR: 1.468660 (0.415519) 0.190240 14.025616\n",
      "RFR: 1.468660 (0.415519) 0.089916 13.834323\n",
      "ETR: 1.468660 (0.415519) 0.000000 13.955045\n"
     ]
    }
   ],
   "source": [
    "##  scoring_mae = \"mean_absolute_error\"\n",
    "data = trainAndEval\n",
    "# Transform in case we use lasso\n",
    "scaler = StandardScaler()\n",
    "X_train, X_test, y_train, y_test = train_test_split(data.drop(\"next_close\",axis=1), data[\"next_close\"], test_size=0.4, shuffle=False)\n",
    "X_train_fit = pd.DataFrame(scaler.fit_transform(X_train),columns = X_train.columns)\n",
    "X_test_fit = pd.DataFrame(  scaler.transform(X_test),columns = X_test.columns)\n",
    "\n",
    "\n",
    "names = []\n",
    "kfold_results_mae = []\n",
    "test_results_mae = []\n",
    "train_results_mae = []\n",
    "\n",
    "for name, model,err in models:\n",
    "    names.append(name)\n",
    "    ## k-fold analysis:\n",
    "    # kfold = KFold(n_splits=num_folds, random_state=seed)\n",
    "    kfold = KFold(n_splits=num_folds)\n",
    "    #converted mean squared error to positive. The lower the better\n",
    "    cv_results_mae = -1* cross_val_score(model, X_train_fit, y_train, cv=kfold, scoring='neg_mean_absolute_error')\n",
    "    kfold_results_mae.append(cv_results_mae)\n",
    "    # Full Training period\n",
    "    model.fit(X_train_fit, y_train)\n",
    "    train_result_mae = mean_absolute_error(model.predict(X_train_fit), y_train)\n",
    "    train_results_mae.append(train_result_mae)\n",
    "    # Test results\n",
    "    test_result_mae = mean_absolute_error(model.predict(X_test_fit), y_test)\n",
    "    test_results_mae.append(test_result_mae)\n",
    "    msg = \"%s: %f (%f) %f %f\" % (name, cv_results.mean(), cv_results.std(), train_result_mae, test_result_mae)\n",
    "    print(msg)\n",
    "res = pd.DataFrame({\"name\" :[x[0] for x in models], \"test_results_mae\":test_results_mae})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "egJPG2UfkL1z"
   },
   "outputs": [],
   "source": [
    "minModelmaeName = list(res[res[\"test_results_mae\"]==min(res[\"test_results_mae\"])][\"name\"])[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "qMFMdtaSPLuc"
   },
   "outputs": [],
   "source": [
    "# Prediction on real test\n",
    "predictionsmae = []\n",
    "for name, model,err in models:\n",
    "  if name == minModelmaeName:\n",
    "    predictionsmae = model.predict(pd.DataFrame(scaler.transform(used_data_xtest),columns = used_data_xtest.columns))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 540
    },
    "id": "wYae4DWlt1bf",
    "outputId": "52b81b8f-d2e8-4711-c0ba-186e879ca09b"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = pyplot.figure()\n",
    "fig.suptitle('Algorithm Comparison: Kfold results')\n",
    "ax = fig.add_subplot(111)\n",
    "pyplot.boxplot(kfold_results_mae)\n",
    "ax.set_xticklabels(names)\n",
    "fig.set_size_inches(15,8)\n",
    "pyplot.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "uY-n8nLzltzg"
   },
   "source": [
    "scoring_ape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "jNlYfyHcl2Zg",
    "outputId": "4f687511-6530-43b2-8b65-ea6d128403e1"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LR: 0.012866 (0.002303) 0.011580 0.011863\n",
      "GBR: 0.031634 (0.005406) 0.009782 0.523425\n",
      "R1: 0.012852 (0.001962) 0.011586 0.011843\n",
      "R0.5: 0.012837 (0.002118) 0.011583 0.011852\n",
      "R0.2: 0.012849 (0.002224) 0.011581 0.011858\n",
      "R3: 0.013137 (0.001505) 0.011603 0.011822\n",
      "LASSO: 0.052573 (0.025804) 0.041304 0.213540\n",
      "EN: 0.061506 (0.021399) 0.043372 0.220679\n",
      "KNN: 0.075191 (0.005131) 0.013003 0.580099\n",
      "CART: 0.035808 (0.003303) 0.000000 0.510909\n",
      "SVR: 0.056035 (0.027715) 0.012300 0.960355\n",
      "ABR: 0.030845 (0.004608) 0.017256 0.557786\n",
      "GBR: 0.031560 (0.005317) 0.009782 0.529330\n",
      "RFR: 0.021510 (0.007971) 0.004667 0.522926\n",
      "ETR: 0.048372 (0.004209) 0.000000 0.525826\n"
     ]
    }
   ],
   "source": [
    "##  scoring_ape = \"neg_mean_absolute_percentage_error\"\n",
    "data = trainAndEval\n",
    "# Transform in case we use lasso\n",
    "scaler = StandardScaler()\n",
    "X_train, X_test, y_train, y_test = train_test_split(data.drop(\"next_close\",axis=1), data[\"next_close\"], test_size=0.4, shuffle=False)\n",
    "X_train_fit = pd.DataFrame(scaler.fit_transform(X_train),columns = X_train.columns)\n",
    "X_test_fit = pd.DataFrame(scaler.transform(X_test),columns = X_test.columns)\n",
    "\n",
    "\n",
    "names = []\n",
    "kfold_results = []\n",
    "test_resultsape = []\n",
    "train_resultsape = []\n",
    "\n",
    "for name, model,err in models:\n",
    "    names.append(name)\n",
    "    ## k-fold analysis:\n",
    "    # kfold = KFold(n_splits=num_folds, random_state=seed)\n",
    "    kfold = KFold(n_splits=num_folds)\n",
    "    #converted mean squared error to positive. The lower the better\n",
    "    cv_results = -1* cross_val_score(model, X_train_fit, y_train, cv=kfold, scoring='neg_mean_absolute_percentage_error')\n",
    "    kfold_results.append(cv_results)\n",
    "    # Full Training period\n",
    "    model.fit(X_train_fit, y_train)\n",
    "    train_result_ape = mean_absolute_percentage_error(model.predict(X_train_fit), y_train)\n",
    "    train_resultsape.append(train_result_ape)\n",
    "    # Test results\n",
    "    test_result_ape = mean_absolute_percentage_error(model.predict(X_test_fit), y_test)\n",
    "    test_resultsape.append(test_result_ape)\n",
    "    msg = \"%s: %f (%f) %f %f\" % (name, cv_results.mean(), cv_results.std(), train_result_ape, test_result_ape)\n",
    "    print(msg)\n",
    "res = pd.DataFrame({\"name\" :[x[0] for x in models], \"test_resultsape\":test_resultsape})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "fi0DaMCaow9E"
   },
   "outputs": [],
   "source": [
    "minModelapeName = list(res[res[\"test_resultsape\"]==min(res[\"test_resultsape\"])][\"name\"])[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "VpRzSQl3piWR"
   },
   "outputs": [],
   "source": [
    "# Prediction on real test\n",
    "predictionsape = []\n",
    "for name, model,err in models:\n",
    "  if name == minModelapeName:\n",
    "    predictionsape = model.predict(pd.DataFrame(scaler.transform(used_data_xtest),columns = used_data_xtest.columns))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 540
    },
    "id": "m96gLBeXt2oA",
    "outputId": "ea7b5145-d2d1-4b78-ed81-a1d343c46ac5"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = pyplot.figure()\n",
    "fig.suptitle('Algorithm Comparison: Kfold results')\n",
    "ax = fig.add_subplot(111)\n",
    "pyplot.boxplot(kfold_results)\n",
    "ax.set_xticklabels(names)\n",
    "fig.set_size_inches(15,8)\n",
    "pyplot.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "oWMVHBVuWJ_d"
   },
   "source": [
    "## Graph prediction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "sJoIfC6BlTXz"
   },
   "source": [
    "mse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 265
    },
    "id": "JPrjxkU5UQFf",
    "outputId": "92c27c7e-7d4a-4569-f324-668e84be3ce7"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "predictionsmse\n",
    "for i in range(1,len(predictionsmse)):\n",
    "  if predictionsmse[i] < 0:\n",
    "    predictionsmse[i] = 0\n",
    "plt.plot(used_data_ytest.index, predictionsmse, label = \"predicted\")\n",
    "plt.plot(used_data_ytest.index, used_data_ytest, label = \"real\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "76RutGTclfxX"
   },
   "source": [
    "mae"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 265
    },
    "id": "kz38DjWclhCw",
    "outputId": "965a8a1a-09d4-459d-d364-21768fb69940"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "predictionsmae\n",
    "for i in range(1,len(predictionsmae)):\n",
    "  if predictionsmae[i] < 0:\n",
    "    predictionsmae[i] = 0\n",
    "plt.plot(used_data_ytest.index, predictionsmae, label = \"predicted\")\n",
    "plt.plot(used_data_ytest.index, used_data_ytest, label = \"real\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "xo2XYFQfCGjj"
   },
   "source": [
    "ape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 265
    },
    "id": "E43amnWfpX8j",
    "outputId": "af1623bc-238b-412c-a2d9-3dea4897f4f7"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "predictionsape\n",
    "for i in range(1,len(predictionsape)):\n",
    "  if predictionsmae[i] < 0:\n",
    "    predictionsmae[i] = 0\n",
    "plt.plot(used_data_ytest.index, predictionsape, label = \"predicted\")\n",
    "plt.plot(used_data_ytest.index, used_data_ytest, label = \"real\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "T3TRusk8WOSW"
   },
   "source": [
    "# Multiple Stock"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2TtuDZaXYdxJ"
   },
   "source": [
    "Objective is to for each stock,\n",
    "- get stock name \n",
    "- best model for that stock \n",
    "- error from prediction on test set at second separation\n",
    "- prediction price \n",
    "- real thing (the test set from first separation)\n",
    "\n",
    "Then select 5 with lowest error from prediction on test set at second separation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "HMGaIGyus-Ez"
   },
   "source": [
    "## Params & models\n",
    "\n",
    "Modify values here"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "0twnD74ZtB_7"
   },
   "outputs": [],
   "source": [
    "import math\n",
    "# params change this to modify the whole thing\n",
    "\n",
    "# For crossvalidation stuff\n",
    "num_folds = 2\n",
    "scoring_mse = \"neg_mean_squared_error\"\n",
    "\n",
    "# For simple trade strategy\n",
    "# Number of stocks to pick minimum error from record\n",
    "nstocks = 5\n",
    "test_size = 0.3\n",
    "\n",
    "\n",
    "# Set model\n",
    "models = []\n",
    "#Do modelName, modelFunc, errorValueTemp\n",
    "models.append(('Linear Regression', LinearRegression(),np.inf))\n",
    "# Boosting methods\n",
    "models.append(('Gradient Boosting Regressor', GradientBoostingRegressor(),np.inf))\n",
    "\n",
    "# Normally ridge would be one model but since it's a model that run fastest\n",
    "models.append(('R1',Ridge(alpha=1),np.inf))\n",
    "models.append(('R0.5',Ridge(alpha=0.5),np.inf))\n",
    "models.append(('R0.2',Ridge(alpha=0.2),np.inf))\n",
    "models.append(('R3',Ridge(alpha=3),np.inf))\n",
    "models.append(('LASSO', Lasso(),np.inf))\n",
    "models.append(('EN',ElasticNet(),np.inf))\n",
    "models.append(('KNN',KNeighborsRegressor(),np.inf))\n",
    "models.append(('CART',DecisionTreeRegressor(),np.inf))\n",
    "models.append(('SVR',SVR(),np.inf))\n",
    "models.append(('MLP',MLPRegressor(),np.inf))\n",
    "# Boosting methods\n",
    "models.append(('ABR',AdaBoostRegressor(),np.inf))\n",
    "models.append(('GBR',GradientBoostingRegressor(),np.inf))\n",
    "# Bagging methods\n",
    "models.append(('RFR',RandomForestRegressor(),np.inf))\n",
    "models.append(('ETR',ExtraTreesRegressor(),np.inf))\n",
    "\n",
    "# Can add Arima here\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Y7Ts4wNkW8Ol",
    "outputId": "f28c1773-cf07-4a53-8926-8486b082bdbe"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/neural_network/_multilayer_perceptron.py:692: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "# Record for all stocks\n",
    "record = {}\n",
    "\n",
    "for stock in stk_tickers:\n",
    "  # Select a stock\n",
    "  pred_stock = stock\n",
    "  # Remove the stock out from concatinating\n",
    "  interested_stock = stk_tickers.copy()\n",
    "  interested_stock.remove(pred_stock)\n",
    "  interested_stock\n",
    "\n",
    "  # Set stock dataframe\n",
    "  temp = pd.DataFrame({\"close\": stk_data.loc[:, ('Adj Close', pred_stock)],\n",
    "  \"open\": stk_data.loc[:, ('Open', pred_stock)],\n",
    "  \"high\" : stk_data.loc[:, ('High', pred_stock)],\n",
    "  \"low\" : stk_data.loc[:, ('Low', pred_stock)],\n",
    "  \"volume\" : stk_data.loc[:, ('Volume', pred_stock)]})\n",
    "  pred_stockDataFrame = StockDataFrame(temp.copy())\n",
    "\n",
    "  for stock in interested_stock:\n",
    "    temp[stock] = pd.DataFrame({stock: stk_data.loc[:, ('Adj Close', stock)]})\n",
    "    \n",
    "  temp['next_close'] = temp['close'].shift(-1)\n",
    "  # Add a feature\n",
    "  temp[\"movement percent\"] = np.where((temp[\"close\"]-temp[\"open\"])==0 , 0,  (temp[\"high\"]-temp[\"low\"])/(temp[\"close\"]-temp[\"open\"]))\n",
    "\n",
    "  temp.drop([\"open\",\"high\",\"low\"],axis=1,inplace=True)\n",
    "\n",
    "  used_data = temp.dropna().copy()\n",
    "\n",
    "  # First separation, one for training and choosing, the other for testing\n",
    "  used_data_xtrain, used_data_xtest, used_data_ytrain, used_data_ytest = train_test_split(used_data.drop(\"next_close\",axis=1), used_data[\"next_close\"], test_size=test_size, shuffle=False)\n",
    "  trainAndEval = pd.concat([used_data_xtrain,used_data_ytrain],axis=1) \n",
    "\n",
    "  # Performing the transformation\n",
    "  data = trainAndEval\n",
    "  # In case we use lasso\n",
    "  scaler = StandardScaler()\n",
    "  X_train, X_test, y_train, y_test = train_test_split(data.drop(\"next_close\",axis=1), data[\"next_close\"], test_size=0.4, shuffle=False)\n",
    "  X_train_fit = pd.DataFrame(scaler.fit_transform(X_train),columns = X_train.columns)\n",
    "  X_test_fit = pd.DataFrame(scaler.transform(X_test),columns = X_test.columns)\n",
    "\n",
    "\n",
    "  names = []\n",
    "  kfold_results = []\n",
    "  test_results = []\n",
    "  train_results = []\n",
    "\n",
    "  for name, model,err in models:\n",
    "      names.append(name)\n",
    "      ## k-fold analysis:\n",
    "      # kfold = KFold(n_splits=num_folds, random_state=seed)\n",
    "      kfold = KFold(n_splits=num_folds)\n",
    "      # converted mean squared error to positive. The lower the better\n",
    "      cv_results = -1* cross_val_score(model, X_train_fit, y_train, cv=kfold, scoring=scoring_mse)\n",
    "      kfold_results.append(cv_results)\n",
    "      # Full Training period\n",
    "      model.fit(X_train_fit, y_train)\n",
    "      train_result = mean_squared_error(model.predict(X_train_fit), y_train)\n",
    "      train_results.append(train_result)\n",
    "      # Test results\n",
    "      test_result = mean_squared_error(model.predict(X_test_fit), y_test)\n",
    "      test_results.append(test_result)\n",
    "  res = pd.DataFrame({\"name\" :[x[0] for x in models], \"test_results\":test_results})\n",
    "  # Get the minimum mean squared error model\n",
    "  minModelName = list(res[res[\"test_results\"]==min(res[\"test_results\"])][\"name\"])[0]\n",
    "  \n",
    "  # Prediction on real test\n",
    "  predictions = []\n",
    "  record_model = None\n",
    "  for name, model,err in models:\n",
    "    if name == minModelName:\n",
    "      predictions = model.predict(pd.DataFrame(scaler.transform(used_data_xtest),columns = used_data_xtest.columns))\n",
    "      \n",
    "      record_model = model\n",
    "  for i in range(1,len(predictions)):\n",
    "    if predictions[i] < 0:\n",
    "      predictions[i] = 0\n",
    "  record[pred_stock] = {\"model_name\":minModelName, \n",
    "                        \"model_func\":record_model,\n",
    "                        \"error\":min(res[\"test_results\"]),\n",
    "                        \"pred\":pd.DataFrame(predictions,index=used_data_ytest.index),\n",
    "                        \"real\":used_data_ytest}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "BVqJFxnXt6bg"
   },
   "source": [
    "## Retrieving stocks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "oY2F3rOYsvxs"
   },
   "outputs": [],
   "source": [
    "# Sorted dict based on minimum error\n",
    "# [0: nstocks] means get n number of stocks\n",
    "min_err_list = {}\n",
    "if nstocks < 0:\n",
    "  raise ValueError(\"nstocks must be more than 0, minimum is 1\") \n",
    "elif nstocks > len(record):\n",
    "  min_err_list = record\n",
    "else:\n",
    "  min_err_list=sorted(record.items(),key=lambda x: x[1]['error'],reverse=False)[0:nstocks]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "f4ff2kPkt9ww",
    "outputId": "78f80807-c0dc-4217-cd6d-d8776f3742b2"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('CSCO',\n",
       "  {'model_name': 'Linear Regression',\n",
       "   'model_func': LinearRegression(),\n",
       "   'error': 0.07931467562398557,\n",
       "   'pred':                     0\n",
       "   Date                 \n",
       "   2017-08-07  26.641494\n",
       "   2017-08-08  26.498473\n",
       "   2017-08-09  26.457270\n",
       "   2017-08-10  25.950669\n",
       "   2017-08-11  26.339263\n",
       "   ...               ...\n",
       "   2023-02-21  48.003259\n",
       "   2023-02-22  47.709703\n",
       "   2023-02-23  47.804224\n",
       "   2023-02-24  46.964166\n",
       "   2023-02-27  47.380045\n",
       "   \n",
       "   [1399 rows x 1 columns],\n",
       "   'real': Date\n",
       "   2017-08-07    26.773432\n",
       "   2017-08-08    26.731161\n",
       "   2017-08-09    26.207026\n",
       "   2017-08-10    26.604353\n",
       "   2017-08-11    26.917150\n",
       "                   ...    \n",
       "   2023-02-21    49.310001\n",
       "   2023-02-22    49.209999\n",
       "   2023-02-23    48.480000\n",
       "   2023-02-24    48.730000\n",
       "   2023-02-27    48.419998\n",
       "   Name: next_close, Length: 1399, dtype: float64}),\n",
       " ('AAPL',\n",
       "  {'model_name': 'Linear Regression',\n",
       "   'model_func': LinearRegression(),\n",
       "   'error': 0.12365379211257639,\n",
       "   'pred':                      0\n",
       "   Date                  \n",
       "   2017-08-07   37.218747\n",
       "   2017-08-08   37.517295\n",
       "   2017-08-09   37.748613\n",
       "   2017-08-10   36.550903\n",
       "   2017-08-11   37.056270\n",
       "   ...                ...\n",
       "   2023-02-21  147.472785\n",
       "   2023-02-22  147.235900\n",
       "   2023-02-23  148.408180\n",
       "   2023-02-24  145.737950\n",
       "   2023-02-27  146.828872\n",
       "   \n",
       "   [1399 rows x 1 columns],\n",
       "   'real': Date\n",
       "   2017-08-07     37.712788\n",
       "   2017-08-08     37.943661\n",
       "   2017-08-09     36.735100\n",
       "   2017-08-10     37.245949\n",
       "   2017-08-11     37.806492\n",
       "                    ...    \n",
       "   2023-02-21    148.910004\n",
       "   2023-02-22    149.399994\n",
       "   2023-02-23    146.710007\n",
       "   2023-02-24    147.919998\n",
       "   2023-02-27    147.410004\n",
       "   Name: next_close, Length: 1399, dtype: float64}),\n",
       " ('GOOGL',\n",
       "  {'model_name': 'Linear Regression',\n",
       "   'model_func': LinearRegression(),\n",
       "   'error': 0.177802053806566,\n",
       "   'pred':                     0\n",
       "   Date                 \n",
       "   2017-08-07  47.341529\n",
       "   2017-08-08  47.271587\n",
       "   2017-08-09  47.055796\n",
       "   2017-08-10  46.225686\n",
       "   2017-08-11  46.550163\n",
       "   ...               ...\n",
       "   2023-02-21  90.946023\n",
       "   2023-02-22  90.833034\n",
       "   2023-02-23  90.051124\n",
       "   2023-02-24  88.316600\n",
       "   2023-02-27  89.048253\n",
       "   \n",
       "   [1399 rows x 1 columns],\n",
       "   'real': Date\n",
       "   2017-08-07    47.209499\n",
       "   2017-08-08    47.004002\n",
       "   2017-08-09    46.179501\n",
       "   2017-08-10    46.504501\n",
       "   2017-08-11    46.946499\n",
       "                   ...    \n",
       "   2023-02-21    91.650002\n",
       "   2023-02-22    90.889999\n",
       "   2023-02-23    89.129997\n",
       "   2023-02-24    89.870003\n",
       "   2023-02-27    90.059998\n",
       "   Name: next_close, Length: 1399, dtype: float64}),\n",
       " ('AMZN',\n",
       "  {'model_name': 'Linear Regression',\n",
       "   'model_func': LinearRegression(),\n",
       "   'error': 0.22810571480116726,\n",
       "   'pred':                     0\n",
       "   Date                 \n",
       "   2017-08-07  49.242454\n",
       "   2017-08-08  49.124322\n",
       "   2017-08-09  48.740716\n",
       "   2017-08-10  47.489207\n",
       "   2017-08-11  48.038026\n",
       "   ...               ...\n",
       "   2023-02-21  94.082565\n",
       "   2023-02-22  95.287457\n",
       "   2023-02-23  95.312748\n",
       "   2023-02-24  93.016107\n",
       "   2023-02-27  93.274087\n",
       "   \n",
       "   [1399 rows x 1 columns],\n",
       "   'real': Date\n",
       "   2017-08-07    49.492001\n",
       "   2017-08-08    49.100498\n",
       "   2017-08-09    47.846001\n",
       "   2017-08-10    48.399502\n",
       "   2017-08-11    49.165001\n",
       "                   ...    \n",
       "   2023-02-21    95.790001\n",
       "   2023-02-22    95.820000\n",
       "   2023-02-23    93.500000\n",
       "   2023-02-24    93.760002\n",
       "   2023-02-27    94.230003\n",
       "   Name: next_close, Length: 1399, dtype: float64}),\n",
       " ('MSFT',\n",
       "  {'model_name': 'R3',\n",
       "   'model_func': Ridge(alpha=3),\n",
       "   'error': 0.45070955754494896,\n",
       "   'pred':                      0\n",
       "   Date                  \n",
       "   2017-08-07   66.529491\n",
       "   2017-08-08   66.770603\n",
       "   2017-08-09   66.612497\n",
       "   2017-08-10   65.606586\n",
       "   2017-08-11   66.491173\n",
       "   ...                ...\n",
       "   2023-02-21  246.539020\n",
       "   2023-02-22  245.754904\n",
       "   2023-02-23  245.197867\n",
       "   2023-02-24  242.146291\n",
       "   2023-02-27  244.460095\n",
       "   \n",
       "   [1399 rows x 1 columns],\n",
       "   'real': Date\n",
       "   2017-08-07     67.673927\n",
       "   2017-08-08     67.376404\n",
       "   2017-08-09     66.390923\n",
       "   2017-08-10     67.404320\n",
       "   2017-08-11     68.417717\n",
       "                    ...    \n",
       "   2023-02-21    251.509995\n",
       "   2023-02-22    254.770004\n",
       "   2023-02-23    249.220001\n",
       "   2023-02-24    250.160004\n",
       "   2023-02-27    249.419998\n",
       "   Name: next_close, Length: 1399, dtype: float64})]"
      ]
     },
     "execution_count": 325,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "min_err_list"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Gr6lRKw8uL1A"
   },
   "source": [
    "# A bunch of trade strategy\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "-ph_InQb2PTO"
   },
   "source": [
    "### trade params\n",
    "\n",
    "Modify here to change starting value and result of trade strategy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "zetZ8xqS2Qxg"
   },
   "outputs": [],
   "source": [
    "#starting cash\n",
    "cash = 10000\n",
    "#transaction cost\n",
    "transac = 0.002\n",
    "#days to hold\n",
    "hold_days = 5\n",
    "#new params for rebalance strategy\n",
    "rebal_days = 20"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "G2gCW2GdyZKh"
   },
   "source": [
    "### Equally weighted\n",
    "\n",
    "When more than half of the prediction in n number nstocks goes up, we buy all of them equally. \n",
    "We start with cash of 10,000. \n",
    "So if we have two stocks at price 70 and 80, We buy 10000/2/70 and 10000/2/80 amount.\n",
    "\n",
    "We then hold it for 5 days (1 week of trading month)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "FQSuxzZrOFgE"
   },
   "source": [
    "#### Set up"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "vlxGvFtGxEYC"
   },
   "outputs": [],
   "source": [
    "n_decision = math.ceil(nstocks/2.0)\n",
    "cash = 10000\n",
    "stock_pos = [0]*nstocks\n",
    "starting_cash = cash \n",
    "hold_days = 5\n",
    "days = hold_days\n",
    "#transaction cost\n",
    "transac = 0.002\n",
    "#Cash money\n",
    "money_record = [0]* len(min_err_list[0][1][\"real\"].index)\n",
    "holding = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "Q56T4AuJuVeV"
   },
   "outputs": [],
   "source": [
    "idx = 0\n",
    "for i in range(0,len(min_err_list[0][1][\"real\"].index)):\n",
    "  pred = np.array([0.0]*nstocks)\n",
    "  real = np.array([0.0]*nstocks)\n",
    "  for j in range(0,len(min_err_list)):\n",
    "    stockInfo = min_err_list[j][1]\n",
    "    pred[j] = stockInfo[\"pred\"].iloc[i,0]\n",
    "\n",
    "    real[j] = stockInfo[\"real\"][i]\n",
    "  # number where prediction up i\n",
    "  pred_up = (sum(pred>real))\n",
    "  if (not holding) & (pred_up >=n_decision):\n",
    "    holding = True \n",
    "    # position array\n",
    "    stock_pos = np.array([starting_cash*(1-transac)]*nstocks)/nstocks / real\n",
    "    money_record[i] = sum(stock_pos*real)\n",
    "    # print(\"buy\")\n",
    "    # print(sum(stock_pos*real))\n",
    "    \n",
    "  elif holding:\n",
    "    # value of portfolio\n",
    "    # print(\"hold\")\n",
    "    money_record[i] = sum(stock_pos*real)\n",
    "    days = days-1\n",
    "    # if days to hold is 0, we sell\n",
    "    if days == 0 :\n",
    "      #print(\"sell and cash\")\n",
    "      holding = False\n",
    "      days = hold_days\n",
    "\n",
    "      starting_cash = sum(stock_pos*real)*(1-transac)\n",
    "      #print(starting_cash)\n",
    "      stock_pos = [0]*nstocks\n",
    "  else :\n",
    "    money_record[i] = starting_cash\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "VZLIGBSABqLk"
   },
   "source": [
    "#### Result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 282
    },
    "id": "xkM0UdlCxzuc",
    "outputId": "1562c3a7-2a71-49f2-a493-77b5e7606f59"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f291e01d220>"
      ]
     },
     "execution_count": 329,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plt.plot(money_record, label=\"cash\")\n",
    "profit = (np.array(money_record)-cash).tolist()\n",
    "plt.plot(profit, label=\"profit\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4qOTDaxj8j3X"
   },
   "source": [
    "### Indicator on one stock\n",
    "\n",
    "Choose the minimum error stock, throw in a bunch of indicator and trade according to indicator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "id": "_6W1JuZ89Cal",
    "outputId": "50593aaf-a8f5-4941-de69-1e5bbf77e528"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "type": "string"
      },
      "text/plain": [
       "'CSCO'"
      ]
     },
     "execution_count": 330,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# First minimum error stock\n",
    "min_err_list[0][0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "JG2-rQZ8_1r8"
   },
   "source": [
    "#### Setting up"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "2yl7Cp-f8mYW"
   },
   "outputs": [],
   "source": [
    "# Record for all stocks\n",
    "record = {}\n",
    "# Select a stock\n",
    "pred_stock = min_err_list[0][0]\n",
    "# Remove the stock out from concatinating\n",
    "interested_stock = stk_tickers.copy()\n",
    "interested_stock.remove(pred_stock)\n",
    "interested_stock\n",
    "\n",
    "# Set stock dataframe\n",
    "temp = pd.DataFrame({\"close\": stk_data.loc[:, ('Adj Close', pred_stock)],\n",
    "\"open\": stk_data.loc[:, ('Open', pred_stock)],\n",
    "\"high\" : stk_data.loc[:, ('High', pred_stock)],\n",
    "\"low\" : stk_data.loc[:, ('Low', pred_stock)],\n",
    "\"volume\" : stk_data.loc[:, ('Volume', pred_stock)]})\n",
    "pred_stockDataFrame = StockDataFrame(temp.copy())\n",
    "\n",
    "for stock in interested_stock:\n",
    "  temp[stock] = pd.DataFrame({stock: stk_data.loc[:, ('Adj Close', stock)]})\n",
    "  \n",
    "temp['next_close'] = temp['close'].shift(-1)\n",
    "# Add a feature\n",
    "temp[\"movement percent\"] = np.where((temp[\"close\"]-temp[\"open\"])==0 , 0,  (temp[\"high\"]-temp[\"low\"])/(temp[\"close\"]-temp[\"open\"]))\n",
    "\n",
    "\n",
    "used_data = temp.dropna().copy()\n",
    "\n",
    "temp = pd.DataFrame({\"close\": used_data[\"close\"],\n",
    "\"open\": used_data[\"open\"],\n",
    "\"high\" : used_data[\"high\"],\n",
    "\"low\" : used_data[\"low\"],\n",
    "\"volume\" : used_data[\"volume\"]})\n",
    "pred_stockDataFrame = StockDataFrame(temp.copy())\n",
    "pred_stockDataFrame.dropna(inplace=True)\n",
    "\n",
    "indicatorFrame = pd.DataFrame()\n",
    "# In MACD the longer it stays above 0 the better\n",
    "# gt is greather than\n",
    "a = pred_stockDataFrame['macd'].gt(0)\n",
    "b = a.cumsum()\n",
    "indicatorFrame['macd_over_0_days'] = b-b.mask(a).ffill().fillna(0).astype(int)\n",
    "\n",
    "\n",
    "# RSI more than 70 is good less than 30 is bad\n",
    "indicatorFrame[\"RSI_70\"]=pred_stockDataFrame['rsi_12'].gt(70).astype(int)\n",
    "\n",
    "indicatorFrame[\"RSI_30\"]=pred_stockDataFrame['rsi_12'].lt(70).astype(int)\n",
    "\n",
    "# Buy and sell signal from SMA\n",
    "# Cross up is buy\n",
    "indicatorFrame[\"SMA_cross_up\"] = pred_stockDataFrame['close_50_sma_xd_close_20_sma'].astype(int)\n",
    "\n",
    "# Crossover Sell using sma\n",
    "indicatorFrame[\"SMA_cross_down\"] = pred_stockDataFrame['close_20_sma_xd_close_50_sma'].astype(int)\n",
    "# OBV open balance volume\n",
    "# A rising trend should be accompanied by increasing volume\n",
    "indicatorFrame[\"vol_delta\"] = pred_stockDataFrame['volume_delta'].dropna().astype(int)\n",
    "a = indicatorFrame['vol_delta'].gt(0)\n",
    "b = a.cumsum()\n",
    "indicatorFrame['vol_delta_sum'] = b-b.mask(a).ffill().fillna(0).astype(int)\n",
    "\n",
    "# Bollinger band controls how high the stock should be \n",
    "indicatorFrame['bb_up'] = pred_stockDataFrame['boll_ub']>pred_stockDataFrame[\"high\"].astype(int)\n",
    "\n",
    "indicatorFrame['bb_down'] = pred_stockDataFrame['boll_lb']>pred_stockDataFrame[\"low\"].astype(int)\n",
    "indicatorFrame.dropna(inplace = True)\n",
    "\n",
    "return_period = 1\n",
    "Y = np.log(stk_data.loc[:, ('Adj Close', pred_stock)]).diff(return_period).\\\n",
    "shift(-return_period)\n",
    "#If stock will go up in one day\n",
    "# If difference is 0, convention is to consider it goes up\n",
    "up_data = pd.DataFrame()\n",
    "up_data[\"1_day\"] =temp[\"close\"].diff(1).shift(-1).dropna().ge(0).astype(int)\n",
    "up_data[\"5_day\"] =temp[\"close\"].diff(5).shift(-5).dropna().ge(0).astype(int)\n",
    "\n",
    "up_data.dropna(inplace= True)\n",
    "up_data.head()\n",
    "\n",
    "usedData = pd.concat([up_data,indicatorFrame],axis=1)\n",
    "usedData.dropna(inplace=True)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "esi_ozp-Jdjn"
   },
   "source": [
    "#### Machine learning part"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 331
    },
    "id": "KoV0tEza-1zZ",
    "outputId": "357d7211-38a1-49e7-8e31-f5408317713a"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-19e11a58-0f57-453e-9bd3-ae884495ecfc\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>val</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>macd_over_0_days</td>\n",
       "      <td>0.218984</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>RSI_70</td>\n",
       "      <td>0.006543</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>RSI_30</td>\n",
       "      <td>0.004124</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SMA_cross_up</td>\n",
       "      <td>0.002977</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>SMA_cross_down</td>\n",
       "      <td>0.003405</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>vol_delta</td>\n",
       "      <td>0.736263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>vol_delta_sum</td>\n",
       "      <td>0.027177</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>bb_up</td>\n",
       "      <td>0.000527</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>bb_down</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-19e11a58-0f57-453e-9bd3-ae884495ecfc')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-19e11a58-0f57-453e-9bd3-ae884495ecfc button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-19e11a58-0f57-453e-9bd3-ae884495ecfc');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "               name       val\n",
       "0  macd_over_0_days  0.218984\n",
       "1            RSI_70  0.006543\n",
       "2            RSI_30  0.004124\n",
       "3      SMA_cross_up  0.002977\n",
       "4    SMA_cross_down  0.003405\n",
       "5         vol_delta  0.736263\n",
       "6     vol_delta_sum  0.027177\n",
       "7             bb_up  0.000527\n",
       "8           bb_down  0.000000"
      ]
     },
     "execution_count": 332,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "\n",
    "clf_1_day = DecisionTreeClassifier(random_state=0)\n",
    "X_train, X_test, y_train, y_test = train_test_split(usedData[indicatorFrame.columns], usedData[up_data.columns], test_size=test_size, shuffle=False)\n",
    "\n",
    "clf_1_day.fit(X_train, y_train[\"1_day\"])\n",
    "\n",
    "clf_5_day = DecisionTreeClassifier(random_state=0)\n",
    "\n",
    "clf_5_day.fit(X_train, y_train[\"5_day\"])\n",
    "\n",
    "pd.DataFrame({\"name\":clf_1_day.feature_names_in_, \" val\":clf_1_day.feature_importances_})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Et73HMZJ9TJX",
    "outputId": "bf372131-8ff7-4f9a-d600-4fce9e621d03"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0       0.51      0.47      0.49       709\n",
      "         1.0       0.49      0.53      0.51       689\n",
      "\n",
      "    accuracy                           0.50      1398\n",
      "   macro avg       0.50      0.50      0.50      1398\n",
      "weighted avg       0.50      0.50      0.50      1398\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0       0.47      0.42      0.44       659\n",
      "         1.0       0.53      0.58      0.55       739\n",
      "\n",
      "    accuracy                           0.50      1398\n",
      "   macro avg       0.50      0.50      0.50      1398\n",
      "weighted avg       0.50      0.50      0.50      1398\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "# Basically using a lot of indicators mean we are correct half of the time\n",
    "# From this result\n",
    "# Can try with different time period\n",
    "print(classification_report(clf_1_day.predict(X_test),y_test[\"1_day\"]))\n",
    "print(classification_report(clf_5_day.predict(X_test),y_test[\"5_day\"]))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "XhrHjeAEAUI5"
   },
   "source": [
    "#### Performing trade"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "RUwkKKCz_aMi",
    "outputId": "ac805fd4-908a-4c97-c64f-c7c26062a39c"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-334-e2cd8ecf16b7>:2: FutureWarning: In a future version, object-dtype columns with all-bool values will not be included in reductions with bool_only=True. Explicitly cast to bool dtype instead.\n",
      "  eval_profit = pd.concat([temp,X_test],axis=1).dropna()\n"
     ]
    }
   ],
   "source": [
    "# A very manual way of evaluation profit of the strategy\n",
    "eval_profit = pd.concat([temp,X_test],axis=1).dropna()\n",
    "# Whether we hold a stock\n",
    "hold = False\n",
    "#Net Profit\n",
    "# Let's make it array so we can keep track of profit\n",
    "money_rec_single = [0] * len(eval_profit)\n",
    "pos = 0 \n",
    "starting_cash = cash \n",
    "#Transaction Percent\n",
    "transac_percent = transac\n",
    "# num of days to hold\n",
    "hold_days = hold_days\n",
    "# leftover days\n",
    "days = hold_days\n",
    "idx = 0\n",
    "for index, row in eval_profit.iterrows():\n",
    "  # Obtain prediction\n",
    "  day_1_up_pred = clf_1_day.predict(pd.DataFrame(row[indicatorFrame.columns]).transpose())[0]\n",
    "  day_5_up_pred = clf_5_day.predict(pd.DataFrame(row[indicatorFrame.columns]).transpose())[0]\n",
    "  # If both prediction says up and we are not holding\n",
    "  if (day_1_up_pred>0) & (day_5_up_pred>0) & (not hold):\n",
    "    hold = True\n",
    "    #Let's buy when market close\n",
    "\n",
    "    pos = (1-transac_percent)*starting_cash/row[\"close\"]\n",
    "    money_rec_single[idx] = pos*row[\"close\"]\n",
    "  # If we are holding then we countdown days to sell\n",
    "  elif hold:\n",
    "    days = days -1\n",
    "    money_rec_single[idx] = pos*row[\"close\"]\n",
    "    #if days is 0 we sell\n",
    "    if days ==0:\n",
    "      #Sell stocks\n",
    "      # Reset conditions\n",
    "      starting_cash = pos*row[\"close\"]\n",
    "      days=hold_days\n",
    "      hold = False\n",
    "      pos = 0\n",
    "  else:\n",
    "     money_rec_single[idx] = starting_cash\n",
    "\n",
    "  \n",
    "  idx+=1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "z0G3PsODBv1X"
   },
   "source": [
    "#### Result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 284
    },
    "id": "nBu3iQl5_c8a",
    "outputId": "6e4437ee-a81f-4dec-946b-8e2ded5d2863"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f291efe68b0>"
      ]
     },
     "execution_count": 335,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAD6CAYAAABNu5eFAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/NK7nSAAAACXBIWXMAAAsTAAALEwEAmpwYAABSa0lEQVR4nO2dd5wV1fn/P+e2vVvZwlIXWJoColQBRVFQFEtCTExsiZqY+DXRJGoSa6KJSb5fkl+aSWyJGo0FjcaCXVBQEAXpHVlggaVtYfvuLTNzfn/MnLln5s6te/s979drXztzZubOuXNnnnnOc55CKKUQCAQCQX5gS3cHBAKBQJA6hNAXCASCPEIIfYFAIMgjhNAXCASCPEIIfYFAIMgjhNAXCASCPCJqoU8IsRNCNhJC3tTWRxJC1hBC6gghLxJCXFp7gbZep22v5T7jbq19NyHkwoR/G4FAIBCExRHDvj8GsBNAmbb+OwB/ppS+QAh5FMANAB7R/rdSSscQQq7U9ruCEDIBwJUATgEwBMAyQshJlFI51An79+9Pa2trY/1OAoFAkNesX7++mVJabbUtKqFPCKkBcAmA3wK4nRBCAMwDcLW2y9MAfglV6C/UlgHgZQB/1/ZfCOAFSqkXwH5CSB2AGQA+DXXe2tparFu3LpouCgQCgUCDEHIg1LZozTt/AXAHAEVbrwLQRimVtPUGAEO15aEADgGAtr1d219vtzhGIBAIBCkgotAnhFwKoJFSuj4F/QEh5EZCyDpCyLqmpqZUnFIgEAjyhmg0/dkAvkwIqQfwAlSzzoMAygkhzDxUA+CwtnwYwDAA0Lb3A9DCt1sco0Mp/QeldDqldHp1taVJSiAQCARxEtGmTym9G8DdAEAIORfATyml1xBCXgJwOdQXwXUAXtcOWaKtf6pt/5BSSgkhSwA8Twj5E9SJ3LEA1sbaYb/fj4aGBng8nlgPzWrcbjdqamrgdDrT3RWBQJDFxOK9Y+ZOAC8QQn4DYCOAJ7T2JwA8o03UnoDqsQNK6XZCyH8A7AAgAbg5nOdOKBoaGlBaWora2lqo88O5D6UULS0taGhowMiRI9PdHYFAkMXEJPQppSsArNCW90H1vjHv4wHw9RDH/xaqB1DceDyevBL4AEAIQVVVFcQch0Ag6CtZGZGbTwKfkY/fWSAQJJ6sFPqC7MYryViy+Qhau33p7opAkHcIoZ8GXnrpJYwfPx5z587FunXr8KMf/QgAsGLFCqxevTrNvUs+J//8Xfxo8Ub8cPHGdHdFIMg7+jKRKwiDLMuw2+2W25544gn885//xFlnnQUAmD59OgBV6JeUlODMM89MWT/TyZH23nR3QSCIivrmboyoKsoJM6vQ9OOgvr4e48aNwzXXXIPx48fj8ssvR09PD2pra3HnnXdi6tSpeOmll7B48WKceuqpmDhxIu68804AwAMPPIBVq1bhhhtuwM9+9jOsWLECl156Kerr6/Hoo4/iz3/+MyZPnoyVK1em+VsmH4ct+x8gQe7T0NqDc/+wAiPvfhtH2rJfUclqTf9Xb2zHjiMdCf3MCUPKcP+XTom43+7du/HEE09g9uzZ+M53voOHH34YAFBVVYUNGzbgyJEjmDVrFtavX4+KigpccMEFeO2113Dffffhww8/xB/+8AdMnz4dK1asAKDmGbrppptQUlKCn/70pwn9TpmKwyZ0DkFo/ru+AU+s2o//3HQGSgrSJ6raevz68roDrfhyeWHSztXe60dJgQP2JCpE4qmLk2HDhmH27NkAgG9+85tYtWoVAOCKK64AAHz++ec499xzUV1dDYfDgWuuuQYff/xx2vqbKVBK9WUh8wXheOzjvdhxtANbG9rT2g+/rOjLLV3epJ2nrrELk371Pm5+bgNkhUY+IE6yWtOPRiNPFmbbHlsvLi5OR3eyBv5mbupM3gMkyH7Ki1wAgEOtPTgDVWnrh18O3LPJFMbHO9QsA+9uP4YpD7yPLb9MTskRoWvFycGDB/Hpp2pW6Oeff16flGXMmDEDH330EZqbmyHLMhYvXoxzzjkn7GeWlpais7MzaX3OBCTuoZkzVuRWEoSmzK3qpG096XXt5TV9/gWQaHzceTo8EiRuPZEIoR8nJ598Mh566CGMHz8era2t+P73v2/YPnjwYCxatAhz587FpEmTMG3aNCxcuDDsZ37pS1/Cq6++mtMTubymVOAUt58gNAUO1futlbOppwNeGMtKcgQxAEimF8qtL25Kynmy2ryTThwOB5599llDW319vWH9qquuwlVXXRV0LJu8BYBzzz0X5557LgDgpJNOwpYtWxLd1YyC1/Sf/ewg9jd345O6FpQWOPDJ3fNQ5hYJ5QQqTMNOt6bPC+Nkavp+k2Z/6WlDknIeoWoJUorZJrr7mGrO6vRKeHLV/nR0SZChMAXhRJojt5u5ydtk2vR5oX/xqYOwYOKgpJxHCP04qK2txbZt29LdjaxEMg2Pr5oxXF8+3NqL37y5A7V3vZXqbgkyECYEW7vTa965+5Wt+rKURKHvkwLPhtOePNGclUKfd/vLF3LlO5s1JeahwXhcaPsCDWZWaUqim2SsJGtyFTCajoSfPofb7UZLS0vOCMFoYPn03W53urvSZ8yTVeWFARu+3zBhlj+/r8AaNio8mqZ0Hfubu/H+9mM4raYfAKDIZU+qpv/4qn36sj2J6R6ybiK3pqYGDQ0NeZdbnlXOynYU08u6Hyf0X9t0RF/2ywrsNuvcRYL8gGm+Hr8CWaFJ1X6tWPCXj+GVFHx50hB0eiS09/qDzJOJZF9Tt75sE0I/gNPpFNWjspT9zd3409IvDG39iqy9dfyyArdTCP18hhewkpJ6JcCr2dh7/TLsNgKHjSRtBOrxG4sI2oR5R5ALXP3Pz/DG5iOGtiKXHf/9/hlB+ybTNU6QHfATm2azYCpp7fbBoQn9ZPXj3leNjiFJnMcVQl+QOjp6g70wnHYbpo2oDGo3+ywL8g+Pn9f0Uy/0XZrkberywkYI7HaStPvyvxsaDOvJNO8IoS9IGbLF5HsoOy2v5QnyE68UMHkk02smFMMq1WyaB1p64LATFLsc6PHJEY6Kj4lDywzryfRTEUJfkDKs5sCcWqrNB6+cbGgXmr7A41fgtKtKQTq8uSqLA+7EdhtBcYED3T4pKecqLTDObZlt/IlECH1ByrDS9B3aQz1/wkBDu7DpCzx+Wc+j70+HeccREI8OG0FJgQNdnuQI/bZeP04eWIpff2UiAKBHCH1BLsBra4PK1JgDVj3LXFBFaPr5DaUUXklBsSb05TQoAXxUrJ0JfW9yhH57jw+n1vTDOVrm2YuSlIIByEKXTUFu8JuvTMSa/S2oLi0AEFw60ZckoX+i24dHVtRh2ogKLJg4OCnnEPQd5i4Z0PRTrwTwA1Mm9Pc2dUNRaMJdKtt6/SgvdGJ4VRH2/u/FIiJXkHucP2Eg7r1kgl58xvwQ+ZM0kbtyTxP+uXI/fvDchqR8viAxeDXPHV3TT4N5h48T2HO8C6vqmgEAz605EPdnflLXjP+sO2Roo5Sixyfr3zXZQWhC0xdkJMmy6bPPVSjQ7ZX0B02QWXg0zx1d00+DuY+/Bxu5Km+Nfaj4ds3jawAA35g+TG9j7qhs0jrZCE1fkJEk6yHni2CkK6eLIDLMe6UkjZq+rFCcOTpQppFN7BY4Eis22b2ezMyaPELoCzKSZNn0+SCf5q705mkXhIYFZgU0/TSYd2QFTrsNs0ZV4ntnj8RDV08FANT273sdbD5hpF9imn5qxLEY2wqSSu1db+G6M0bgVwsnxnRcsjR9PoxeEZk8Mxam6afTpu+XKZx2gqe/o6YJqWtUC/4koiteKZBbiik4zgSPIEIhNH1B0mDazNOfxj7xlTShzz2xMqWQZAWH24SZJ9MImHdUwZiOiFxzZk+WGiEeZYFSim7O3bNXi+w90taLJVo+Kpew6QuyHT53SqywIW+ikRVjPpe/fliH2Ys+xBEh+DMK3WXTrWr66ci941cUOEy++kB8o44nVu3HKfe/p6+z73Pvq1vx6zd3ABA2fUEOwIes1zd3Y4Dmkx8NyQqC8ZvMO1sb2gAAWw+3J+V8gvgwm3eSmcc+FLJC4eQ0/b4I/Vc3Hg76bAA43hHwBBI2fUHWww9n/++dnfDJCiYPK8cfvn5axGP3NXcltC/vbz+GVzYcxoQhgcRWskL1Ii7dSXrJCOLDYwrOSkdqZUmmsNssNP04sqGZXxTsJdbOZZ4Vmr4g6/FKxuAWr1/BjJGVGDOg1HL/5787U1/u8SY298iNz6zHu9uPGcwECqW6G57I6pk5ePwyfrR4IwBO6KdlIlcx+M6zEobxaPrmY9h6a0/Ag6zAKYS+IMthpRFLChzo9cvwSnJYH+czx/THyjvmYnhlUVBZxURhSNerBIR+slLmCmKniQt+YuadXUc7Ul4XW1aonhAQCESNx3NvmqvASQqFV5IN911pigIFhdAXJA32bBQ4bOjxyVBo5MCWYZVFsNtIQtzirPBwD5msULjs6sMozDuZAy9Tmab/1w/r8Mxn8ac/iAe/rBgSAfZF0y9yGYW+rFC09xiLCqWqPKgQ+jnAlAfex1+WfRF5xxTDNCK3067bLgsckW9sQuKzm0YD756pUAoK9TxdScqTLogd9tuXFjj0ORcA+GxfS0r7ISnUkAjQxk3k/vy1rXjNNDkbDrPQl2SKDo9R6GeMTZ8Q4iaErCWEbCaEbCeE/EprH0kIWUMIqSOEvEgIcWntBdp6nba9lvusu7X23YSQC5P2rfKM1h4//rJsj77e5ZXw5b+vwoK/fIzGDk/a+sXkNi9oo7Fb2ghJ2lB+2c5GfVmSqT5B+NhH+3CgpTsp5xTEBlMWfnPZRENO++4Ez/OEY/2BVvT4ZD0HEBDIBNvhkfDsZwdx64ubov48sxYvK1Sfp1hwippGeWhFYR97HR3RvFq8AOZRSicBmAxgASFkFoDfAfgzpXQMgFYAN2j73wCgVWv/s7YfCCETAFwJ4BQACwA8TAhJbXn7HMRqqLm/qRtbGtqx61gnVu9NrXbEYyW3XVFoM3ZCLKtsJRqFUkMQ2PYjHck/qSAiLPjJRoghOCpZ8zxW/OG93QCAj75o0ttYX+IppGLuuqQo+rP7lSlDUb/oEt2UlWwiPoFUhfnPObU/CmAegJe19qcBfEVbXqitQ9t+HlHz5y4E8AKl1Esp3Q+gDsCMRHyJfMbK64R3A/t4T1PQ9lTBTCc8LNgmHMk07/B4/ErScvwI4od5fdltJMgskirYBO5lk4fqbSwit8sbeL46TSaaUMiUopS792WF6kLfXEsi2URlRCKE2AkhmwA0AlgKYC+ANkope+U1AGBXZyiAQwCgbW8HUMW3WxzDn+tGQsg6Qsi6pqb0Caxsga+lybweeKHv7UNUbF+xmu8qdTuDG00k07wDAHNOUqsT3b9kux4OL8gcLv3bKgDqfVDkcmBwP7flfu29fvz+3V0Gj6xEIckUk4eV47b5J+ltTNPnzUyt3dEJfUWhGF5ZpNeCljihn+z8+WaiEvqUUplSOhlADVTtfFyyOkQp/QeldDqldHp1dXWyTpMz8L7wf3x/N55bcwA3P68WCCl1O1I6JDZjJbijSUubTO8dAJg2vEJfbuuJ7qEVpB4mC5fefg4AoKzQqDD865P9eHjFXjz32cGEn9sjySh1O/QiP3x/3tp6VG+LdqQoU3VSmJUJlRWqP5uJrsIViZimiymlbQCWAzgDQDkhhI1XagCwqezDAIYBgLa9H4AWvt3iGEGc8FrO4bZe3P/6dn29stiVluyEDHbqH84bo7dFk4rBRpKbVdHNTSa3caOiNL4fBRYwDbikwIEJg8vgNRULZ6PGZEzAe/xKkKcZIQSncBHdQPSJAWWtxCIzG6mavrrNTjJM6BNCqgkh5dpyIYD5AHZCFf6Xa7tdB+B1bXmJtg5t+4dUVfmWALhS8+4ZCWAsgLUJ+h55Sy/3IBzv8GB4VZG+XuxKr6YPzaY/sCwwPB9VXRLxKEJIwvs9qaafvsy7xvH22XTkdxGExsYJw0KXPch7hynIJAlC0+uXUWgxn/DYt6YZ1mMR+nZC9HvPJym6adaWYsf5aE43GMByQsgWAJ8DWEopfRPAnQBuJ4TUQbXZP6Ht/wSAKq39dgB3AQCldDuA/wDYAeBdADdTSoVBtY/wngRfHO/CvqaA1pNsM0kk2LljLQNnt5GEa918xk9+Qu3QiYA7aSpdAgWR4c0eQ8oLcai1x7Cd3SNPra6HolDUNXbiluc39NnGTylFU5cX5YXB809m18tYNX3modPa7cO1T6o6b8Zp+pTSLZTSKZTS0yilEymlD2jt+yilMyilYyilX6eUerV2j7Y+Rtu+j/us31JKR1NKT6aUvpO8r5U/dGqRpNedMcLQvuWXFyTdTBIJ9lA6YlRlEtlvryTj6dX12H28U29TKMW8cQP0deY9sbcpsUneBH2DF4ZD+rmDatPyd8j+lm7c88o2vLnlKDYebOvTedt6/Oj0SBhpUSHLHEDljTJnk6LZ9JnQ/+3bO/VtjhTl0WeIiNwsp1PT9K+cMdzQXuZ2wmZLvJkkFti5Y72pE2neWV/fivuXbDe0nej2o6UrIEBK3A4MLCuIy/9akDz4+U27jYQtXtLpkeDWzDGPr9wXcr9oYJOzVmkRzHEm0ZZxZAVZmMsy72FnyzRNX5DZMEFVUeQK2mZPgm08FuLV9O0kceYdr8Xw2yvJ2NwQyJ9fXuiEw2ZLSyZHQWh4847DRoJiN3jvsI5eP4ZXqhGtB1qMZqBYCbhSBm8zmyr9UWr6MlWFe6HFiyQjXTYFmQsLDuGDnv58xSQA6k2WXvNOfJq+zZa44CxZ08SevWEmtv3qQtw4ZxS+e/Yowz5upx02W2ojPgWR4TVgmzbPE0rb98uBCNfSKAIAwyFzEcFmzAL65fUNUX2momn6hBC8eOMswzah6QtiossrgRCg2GXHHQtOBgDUVqm2SJsNKUlnEAr2eMY6kWtL4AiFae8VxU6UFDhwz8XjUVLgwJTh5Yb97Gl+QQqC4TVtNu8SajQmK1QPTuxrvRXdf95CGJs9heos5oEaOzzY29RlGIlICtU/b+aoKsP+QtMXxESnR0JJgRpEctOc0Xjzh2dhihZ8ZM8Um37ME7mJ8zoK1YfF35ulF22hVL1WQuhnFrxljlWw4u9n/taWFKpP9MZTuNx43ugiZc8fPzBoYtfjlzHz/z7AeX/8yJC3R9X0A/vxufNT/YwKoZ/ldHoklGlBKjYbwcShAX90Gwm2g6aSgE0/Vk2/7w8uQwrxALuddlSWqPMgFFQI/QxE4qQ+E5i8pi+bNGlWkKSvv2O0kbIuBwly2ezxyfp9v+lQG5o6vbjv9W3YfbzToHhM5kaaqb7vhNDPcrq8/pDZ+WwkvMdDsgl478Q4kZvAEYqsBJJ3mSFQ2yhN/wtSEIyfu3eZps8LSH5Z4mz6O4524PvPrsdd/91iyE0VLdFGyjrtNsOLifWD0eWRsHpvM/79qVr8ZefRQBbXh6+Zavk9UoEQ+lnMyj1NeG/7cYMPOk+6g7PYqWO1WRKTeYdStcrQloY2bDvcjl3Hok+BzJ5Bq9EGa1KoWhZPVig2H2rD79/dFVN/BcnBoOlrv1VooU8NEdXvbDuGFz4/hA0HW2M+bzjvHR6n3aa7bC7ZfAQdHr/hRcUnVQNgcCAodTsxVdP2Uy30U5PAWZAUPt9/Iuz29AdnxZdF0Gze+dnLW4K8JOoXXRLVZ4XT9Gv7F+P88QPww3ljcd/r2yArFAsf+gQAcMeCpOUUFEQJ7wNv1yQwL9glk4CVuP2ril1o6fahvrkHZ46O7bzhJnIB4LnvzkR7rx8r9zTDJyvYdrgdP1q8EZdNGYofnzdW388nK7rysvKOuRhWWWT4nJ9dOA63PL8BJw8qja2DfUQI/SymW7NhsnStZhLpBRMP7NSxOieYzTvRusVZEcqmD6ia2uPXnW55TkppUnK6CKJnHCcM2UiN90aTDS8AxfASGNTPjZZuX9T57nmUCMrK7DH9AajlG/2ygsZOtTrdiW6f8aUkK/pnWd1KZ4yuwvpfzI+5f31FmHeymG6vhAGlBVg4OagsAYBM8N5R/zPbebQQk339ytOHhdk7PNF6YthtxKApCvN++nDaCb5/7mjUcmkQmH2dF6q8OX3ZzkbDqLaqpAA2EohYjwXdTz/CPeO02+CXFDy/Rk3tXFLgMIxO/DLVR7up9sUPh9D0s5gurxS2xFqmBGcRAgwtL0RlcXDUsBUEMCRW6YvGHW11IvNErpD56YFSGlSQHAi8tENp+h9/YSy4VOS0o6TAEZemf6RN1dyjmcjt9slBdZcZasCYuiyEviAheEKkf2XYMiTLJiHAJ3fNi/o4QohB6FqVhIyWaLU2h50YqoypL6zMeVDzBZ+sgNLgvDdM6Iey6ZspctlR6LJHnRCNhxUhijQ6dJkKAkmKAj/XPz9n3klx/FVYhHkni5EUGtYd0k7SnVogvqEtgTGvil9WLHOWRIMUg6bPCxGh6acHVr7S/HszAcwL8XDuyAP7ufucTynSszOjttKwLsnUkItH4sw7mTQ/JIR+FiPJwcNgHhshesAK48lV+3G4rTfEEYmF1/RjgRAEafrDKgvx64WnxNyHWGz6oaI9BamDFQUyF0Rnmv8Pntugt4UT6IPK3No8TfyjxEi58s2pPCSFGvrkV6j+DAhNX5AQJEUJK/R3H+9EU6cX72g1PY+1e/DAmzvw3afXpaR/Ae+deDT9wLpPVuBy2HD5tNgndJlmGCkVhMMUkUuFrp8WntECmcxBVWeP7Y+xA0pw6ESPbu4LN19VUuCAw076pOn7pPDHWtV7Zi8Ku43geLsHJ7p9ADLLpi+EfhYjyTRsBssdWgTgh7sa0dDag9te3ATAWCIwmejuajEeZyPEIHT9sgKn3YZClx3fmjUi6glhAGjp8qKiyCm8d1LML5dsx/+9szPyjibe2HIEANDc5TO0u512XDVjuJZuQfXIYUL/pnOCHfGLC+xBL/JwSLIS9KKJpOmbTat8XIzbYcPu45148IM92jYh9AUJQFKoHp5uhe7bTIG/frAHn+5r0dpT87Ozxy1meyYxeml4JUUvXhHrkL2ly4f+JZGLsavRlfxEbvTdFQTz1Op6PPbRvpi9Z648XS0G9O3ZtUHb2MQpr+kPqyzE/AkDgvYtdDlgt9nCFjl5d9sxrD/QioeW1+Gkn7+Dcb9417A92lKIDEKIfr5uk1mVZJCkFd47WYykKHCG0WAdNhv8sgxKKQaWFXLtqdE6aJjAlHCY/fp9kqLnSI81A2e3TzLUGgiFy2EzTBIK805i+Lz+BOaNGxj1/uyeKXUH16dlQp/9TpJWbNxKiSl2MU0/tOC+6dn1QW2bDrXpyyOqgsslhoPA6F3EIzR9QUKQZBrWbMG2Uahh6eb2ZBO3TZ8Ee+8wTV+100avgXn9iqXt1UyBw6aXyQOEph8vslagnBFrvVqfpilb1WAoMAl9mar3P2/iZDUlhlcWxWXT/4qWhqPU7cC0ERUxHUvCpD1JdfHzcAhNP4uRFBqUz5tHF/rU+HLojSPzYDzEa9MnCPbeYVqemjk0+BhKKXr9MopcxlvaK8kotyglacZltxniAYTMj4+Hl9fhj0u/0Ndbun1h9g5GklXnBCuTYIHZvCNTOGw2Q93a/5kzGlfPGI7yIldMNn0z8UTyAiSkOSmDZL7Q9LMZVmw5FExbUqjRvc2qnm4y6JumH1hn3juAda1UAHj2swOYcN97eHPLEezjqhl5peg0fZfDJPSFqh8Xh1qN9Wn5gLdoUGNPrO+XgHlH1ve12YhhQtVuI/pL3jw5n2wopSHnm4R5R5AQ/LIS1nuHN+/wGs/kYeVJ7plKuGRT4SAwee9Iij6isWnam1kov7FZdUu95fmNmPfHj/R2r6QERXdaodr0AyMgIfLjo63HD4eN4H/mjMLgfm7DNY0Gv6zAGcLRoMCh/o7s5axQNU4l1ByVGpwVn5/+vRePj2q/uSdX68sypSHNScJPX5AQZIscJTwOrsQcuxkLnfa4H4RYCXjvxHYcr+k3d3lxpN2ju9Mx26hC1YjM/6w7BK8kG8Lfebx+OUqbvt2Uwz+2PgtU2nr9mDqiAndfPB79Cp1BZpK1+0/g1hc26i/tbq+EZz47oK+Hc0PWvXdkbiLXRkKaOFmNBCvCeYB9Y3oNvjdnVMjtPI98cxquP7MWgPo8Ck1fkFT8cgSXTfbw0ICmX+C0pSwJGwuTj8u8oy2/tvEwAODNLaomz76TrFC8uvEw7nh5Cx5fuT+ke51XUlDgjHybmwWHMO/EBz+y2nWsEx990WS4367/11q8tumI7tJ4/5Lt+MVr23R3YklRQqYWYbZ7ZjLafKgNnR5/yJeEwxZ6ItcXRuibc+qEw+2045dfPgXTR1SoQj/E+TJI5guhn83IimLp5cAIaMWBCj4FDltS7ZzNXV4cbFHtukw7j0bT5iGEBNXXZemV2QtEVijqW7oBqMN9q++0/Ug7Wrp9ulkgHOYBk5D58eGXFLhM92RbT2Ayl90L3V51BHC4VU0JwgZqfpmGdENmL28msNt7/djb1B3SHBTOpv/+9uMhv4PLHnueJ2Z2DC30M0fqC++dLEaKMJEb8N4JTOS6HH1LQmWmpcuLh1fs1W23z36m5havX3QJPJrtNVwmUCv4hGusr3dfpNpYmRK4fHcj2nvVwJ+SAgd2HQsuGXnZQ6ujPqd5NCJkfnz4uUl3xoluH6q0ADn1BexHp0fCwDLAo903bk2gS3JkTZ+fcL/0tMFwOkJo+vbQNv1btej0eeMG4MNdjYZtsWj6+rlsapF0Zt65ff5J+BPnxZRJCE0/i5Hk6Fw2391+DLKWp8fZx8yDZlbsbsITq/bjzS1H8c7WY4ZtTNN3R6Fp8/DmHT1LpqY9MnPWD57boBec7vRau9cxjbC5yxvVOXmEeSc+WMoMAPjd104FYEypwIR7U6f6m/SaIld9cmhvKyaMb31xk57+eMyAkpAR5uFcNtkI8iytCpbhPGFGz6Gwa6Ykdr/eMncMXrhxVsyfkwqE0M9iIrls8rBRgT1ClGKsME3tvVvnBJV+8/jVoKpIuezNEBBuYk9LmMaEvsVHLd0RPFQ3F82OhND0E4OfU0QmaV5iJzhf/ZFaNawnP9kPIKC1M/92rz/0HAxvpvtgp/qb2wgJaeK020L7zU8dXoEzRlXhlCFlQdvi0fTV+BG1Tq+NqOaeYldmGlKE0M9i/BHSMPAcb/fAoQn9cPlIYsWjTaoxbf7muaNht6lC2+OXo5pENcNr+qyvzG5r9ZLbqSWW45n3xxX6cjTeSsKmnxh8nKZfVayadF7bpE7GH2v3YPlutcLV0h3Hce7/W459zeq8DJuIV+MqrEeGvDD+yXw18ra1xxfSXh5O0/crqrtzjalYufk80WLX4kf83ET0BIsXSiYghH6WoigUlCKs9w7Pa5uOaIEsiS2hyGz5TLgXuRyQFQqvpGYtjMZH3gyfWllSFF1zAsJ/34oiJ/7xrWmYNapSt/cD4VPwMsyjEZF7Jz7UlBnqtawuLUB5kRNHtPoNd/53i2HfQ629huMA9X5yhTBZ8mYfNvILV0hFtelbb2fuzkPLC4OquoU6fzjUEbQaJezU79XMmbzlEUI/jdQ3d+O7T6/DL5dsj/lY5pceLjjLrK2qgSyJtekzTZ89kKz4Ra9PVss5xiP0CTH5bQdu03DC+JFvTsMFpwzCCzeegQGlgcyaTOOMdE4eoenHh48LpAOACycMwvYjHbj9xU34yFTHtqIokFSNjeh8YVxseWFs4+I1QhEu4Rrv7jy0vNCwzRmPpk/Uc0VyrsgEhNBPI5/ta8Gyncfx1Op6HO/wxHRstAW/eZx2W8TMg7HildTgJyY0mdDv9knw+BV94i5WePMOb8Lyh6l5yheJZxXDrpg+DL/40oSI5xPmncTgN3nfMFPJK1q8BY/bacdDV08FoGbjBMKnzeBHY2yZT8lhtvKEc9mUwxQgikfTL3CqWVr5iexMJbN7l+PwqXwPtPSE2TMYphnFolU47bbE2/R9xojXQm3y6vGV++GR4jTvEHWkoPo9G4VIOM2umBP6bNj/7bNqDS+DUARP5AqpHw9qEsDAtQxnH7cRgvGDSwEAT6zaj/ZeP3p8ckxxFWxE+Px3Z+Ljn8017OOwEXR6JUtPLHPk74c/OSeqPoei0GlHj3bP8p87Y2RlmKPSgxD6aYT3N+Zt0NHAvFpi0SoOt/XCaU9cRC6lFE9/egAdXKg969dTq+ux/kBrXFoPger+9vPXtmneIIGHKFyxar6uKvuK0XpQCE2/71CqzjPxL9BIAnRYZZFu5tl0qA0HT/RgdHVJxHPZuSA9ADhzTH8MM03KsnxMLJqbRzKlMBlVXYL5E9S8//EEUhW67PD4ZPi1zJ+MZ26YgU33zQ9zZOoRQj+N8MmoYhX6/1nXAADYr3k/WPHlyUOC2pg/cSLwWpha+ELsnR4J6w+0xvy5TLAvXntQS7VrC9pmBa/RM9fAaAPDgmz6UfdWwAgUAeeEfpiXPiGq0vLwNdMABCJ3q0oiZ4GNxqZ/TDOZsolkHqsgMFaox1w2MRqKXHb0+GVtZBr4/gUOe1SpvVNJRKFPCBlGCFlOCNlBCNlOCPmx1l5JCFlKCNmj/a/Q2gkh5K+EkDpCyBZCyFTus67T9t9DCLkueV8rO4hX06eU4plP6wEAO44EuysyfnDuaOx44EJ8enfAO8ERY7nBcLDAmlmjAkPYOWPVrIM/Pm+sfr5Y4YfjB0/0GB6iUA/5up+fbzDvPPLNqfjntdOjKpUIWJh3hKofM0zr5mVpNC677BZhz0O41CKMweVuAIGXezisTIxmTR+A7nQQn9BXvdZe33QkZlNtqolm7CsB+AmldAMhpBTAekLIUgDXA/iAUrqIEHIXgLsA3AngIgBjtb+ZAB4BMJMQUgngfgDToSpS6wkhSyilsauCOYJXqwjlVxS090RfbOL1TUdwpF3VYq7QctJYQQhBkcthKCySSJfNHu3huGzKUL1teFUR6hddAkANkY81BQNgFOxr9p8wbGN9/+qUofjS5CH49r8+B4Ag4V5e5NKH69EgzDvWSLKCX7y+Dd89e1REswsbhfETrtFMirL9WQR1NDWczx5bjee/OxMzR1VF3NfKxGS2vQPAEM2LJ9YIckC13Q8oLUBjZ+To73QT8epSSo9SSjdoy50AdgIYCmAhgKe13Z4G8BVteSGAf1OVzwCUE0IGA7gQwFJK6QlN0C8FsCCRXybT6fFJqL3rLby6UTXNsOjD0gJHTJr+YW64+rVpNTH1IZEum/9dr36PwhB287EDS1FTERz8EgmrIikMpoEP7OfG3JMH4P3b5mDZ7eeE3D9aMin1bSbx7ac+x+K1h/C9f6+LuC8T+nxpQCtPnEVfVdMzsL3Mmn44N2SeM8f0D+vI8M9rpwMImG14fCazIQDcOGcUfnvZxJifKQA4vbYS7982J+bj0kFMccKEkFoAUwCsATCQUspmSI4BYGrVUACHuMMatLZQ7XlDY4eqBfxl2R5cNqVGyzNih9tpM0yGRqLHp+4bT24PewLNOyyhlDsOb4dwhDOtsPcVEywnDSxNyDnNMj/c3EE+sXJPMwBgX1PouSMGG4VZTeTOHlOFT+rU9Mljg34zTdOXIjsnrLxjbtTeNWMGqCMTs9smpRQ9PhnFBUaN3mm34ZqZI6L6bCvKi1y49LTBcZZaTB1RC31CSAmA/wK4lVLawU98UUopISQhTwkh5EYANwLA8OHDE/GRGQPTYJjNkBXtdtltIfPBW9HtlVHmdmBWFEPboD4kcCK32GVHt09OeLh5uDAC3YSQYMU82Kaf2M/PVmaNqsRn+05E5fb62T7VFGcw72gCmv9NWeDclyapjgZsd79u3gn945o9dMLhMJmNth1uxylDyuCVFMgKNcwBJYq/Xz018k5pJqpXJiHECVXgP0cpfUVrPq6ZbaD9Z/lJDwPgDc01WluodgOU0n9QSqdTSqdXV1ebN2c17MZnXi+s9qvNRmLSLHt8Usw37Js/PAvLbj8noTb902rKMaO2Mi4TTjjCXQtmVx6TIA2fIRKuWTNcE7KR7k9ZoboJiLfOsNz0/PHDKouw5p7zcPv8kwAErn00mn4ssBeOJFOs2tOMS/+2Cs98dgBdWlbWTE2Ilmyi8d4hAJ4AsJNS+idu0xIAzAPnOgCvc+3Xal48swC0a2ag9wBcQAip0Dx9LtDa8gaWOqGtx49th9vxxuYjaOvxaSHc0YuZbp9s8EmPholD+2HMgBLYE2TTl2QFn+5riSuhWiRY986wGMlcetpgvHHLWfjypGB31L5gVi7X1Z+w3jHPYApKj08OSoPM09IdmMDk7ezMpk8p8Nnd5+Hze88HAAwsc+tuskzoe+XYbPqRYJq+X1b0gjs7j3bqBVySoelnA9E8sbMBfAvAPELIJu3vYgCLAMwnhOwBcL62DgBvA9gHoA7APwH8AAAopScA/BrA59rfA1pb3sDbFq/XvE5ae/yaph/Yz+OX9epTVvR4JYNHTiwkymXzsY/3AQBW723p82eZYTZ93iuIQQjBqTX9En5Os5/+z17eEmLP3GfXsQ7c8NTnOHSix+BW3OkN7WzAvxCIhU1fphSD+rlRXRrsQst2Z2UQE6XpMz98v6wY6jV3e9W+lhTE7qWTC0SUHJTSVQhMtJs5z2J/CuDmEJ/1JIAnY+lgLsHb7fnCHjZizBb4w8UbsXTHcez934stvRPi0fQZDntibPpH21UPomTU27Vy/Us2GZ4jK6X89YM9+GBXIy48ZZAhAK/HKwMhrGp8UJ6Vn3u4kpm6eUePMk/Mj8HO+Zu3duKnF6imJAI1LxQgNH1BCgglbFkubsaK3er0SG+IIJF4bPqMcDnGY6FCizJM1APKo3vopPDuFC6bAbo0TbjHJxk0/Z4w5h1+G/+iYJ4sVm6TDOY5+eoGdYovUZq+22nHN2epziCbDrXr7cymH+9oOdsRQj+FhDKr2Ew2fabdh7Kh9njj1/TttsQURmd9TEZJuICHjnqOUVFEXfaVKMsS5AXMvNbtk3X3YACGZTO8ds/fX6Oq1d/uoomDQx5LNEMCU3IGlrnj6LU13549EgDQ3hsIfmQ2/Wg8knKR/PzWaSJUdksbCbgIXvTgSj1HPXuQ2nv8eH7tQdx0zigQomYOjNfzwGkneiUpn6TAI8koczsjHBVMp0dCscuOaSMSn0VQD/KxEex8YEFKBHI8SbZyFSa0e30yOj0S+pcUoLnLG1LTp5TiaHsgNTifunv84DJsvu8C9CsKfY+ZTWvRps6IhsBkrvqdCNHMVECQn36+IIR+CmEeBGbsnMmFL/3HHrJ7XtuKt7YcxWk1/TB7TH90eaSww+Vw2LVJY0Wh+M5Tn2NVXbOeNiEWujwSSuLsQySYzLARElcah3joi3nHJylxpePNVJhG/9CKOkO8QihN/y/L9uDBD/bo62YzZjiBDxhfuEtumZ3QIiTsd2XPFwHRzTtC0xcknbtf2WrZbiPEMvUAKyjdoaVokBQKv6yg1y+jNA7tHAhoPpJCsaquOerjlu44jhc/DwRUL9sZXIw8UZjNO6kgXjnz/vZjuPGZ9XjrR2fhlCGJ9ypKNZKs6Hb4m84ZjUdW7NW3hdL095kyvcY6Z8Rf+0EJNO0AAfdP/kXULWz6glQxo1Y1hYwbFHCBYMXKrWp9bjusTj7xUahdUUyMhYOViIv1wXxp3SGs3NOEI229lqlqE0kgRW9ST2Mg3hfMcm3SfTM3UZjNXPbwauxr7salpw3GnQvG4f9dfhoWf0+dtwkl9M1XLlbvMP7am9Md9xU2auDn07p8Elx2W06NzmIhP791mpg0rB+KXHZcziV0stkIbMQ6IpcNQ3lzB5vsitfs4dQ1n8BD8NaWo/q5QuGXFZw0sBRv//hsvP3js/E/54zS3eASDU2Dpt/XM+VKpa2tmqLBsmN+ffowPS6CN+88uGwPbnhKjTVhdSGYUhO7ph+4+on2BnOYlBxm0y/KU3s+IIR+SpEUashACKiJw2w2ApkCq/cazS1MwLMXwqd7W3Dmog8BxO/WFtB8Ag/mzc9vwFOf7I/Ydz5S8u6LxuOWeWPj6kMkmKaYygLT8WqYzB6dc7l6uEtfpOWZZ66cAPDnZV/gg13qKMcnKTh1aD98aZLqoSPFWIOZfyQSXV/Wbgs270hK5texTSb5+83TgKLQoIAjG1FzlSgKxf++vdOwjdke2e26n5sIjndo6rB4CABg57HOsMelsuAzExqpfDCHVhTGdRz7NXOt6EpHb0Crt9kISgsc+tySGZZDas5Jaq4sq0jqcCRT6LP7/eAJNcKdQKuPm8fReELopxCZqjfblOEVetvUERWaRw3FhMHGbJW60NcEisxp5844b1o+NJ3HZ1H6kMdcqzaZ+CX1e6bqfAAwpJ8bdy4Yh3svHh/Tcbng6bl8dyOW7TBOzPN+7QBQVujUhT4fP0Ip1bPFjqgqRv2iS2J24+XNO4ke3Zk/j2hOE6kcRWYa+Tl9nSZkTdOfNqICG38xH/Ut3Rg7sBS3vbgJ3V4JlcVG/+QPdjbi0r+txLbDqhsnr52HKzQSDqZJmQO0rCaSeSRZgSNFLm56OH4KJ9oIIfj+uaPxQZxeSdms57PqY/WLLtFTb5uL+vQrdKK91489xzsx/88f6+3rDrTCJyt9ct9N5tyNlUYvK/kt9IWmn0JkzqZfUezClOEVKClwoLHDg/qWHjz6UcA97ntnj8SMkZUYWBpwYeNtpfGmUmDas99kd430ElE1/dTcLmwUEk2pvUQTq/zRC3QnIQdRMtl2uN1y8p4V8W7rsRb6O7g4EgD4+qOfYktDO8oL43MhBpLrpWUl3PNd6AtNP4XIivVNeLzDWFez1O3ALXPH6kEttXe9BcBogolX6LtCmHcifZ5q00+ReUdOvU2fQWL042F7//KNHbDbbfjWrPgrL6WSS/+2CkP6ubH6bmPOxPIiJ5q7vBg70FgPt1+hE3VNXSHvk4lD449RSGY0tNVnyxZF0fMJoemnEFlRLIW+WQBv/eWFllGMhvwmcWv6mtCXTOadCJq+6r2TKk0/9TZ9nRhPyQuVX7y2LcGdSQ5sjugIlzqB4bAR9C9x4eGrpxna+xU6UdfYFbJsYlWJK+7+pFr+SgrN6wR7QuinEJlaa/q+KPPb89kLzxrTP64+MDu5+ZyRNH2flDpNP9EVlGIhH0QB/1Obf3e/rGDmyKogpWPuuAEAgCWbj1h+JjMLxUMq8x49tboeS3ccT1ihlmxECP0UooSyJUaptDd3eVFe5ET9okswpDw+F8NQ5p1IrtWSosCZolSUv7/8NEwaVo5B/RIbkh8NsQqgbFQY+VGd2afeHI/BWDBxEKpLC3TXRzOVfRD66bC0mONl8gkh9FOAxy+j9q638NbWo5Y3mztCdO2/rj8dgGr77+uEocsRKCHHE2kiV5KthUEymD2mP16/eXZaNP1YBRA/B1BZHL/gSyW8dm/24lJ92K2vO3/v3XfpBMO2igzW9K3mWfJ5IlcI/RTAElgB1tWgCp3hhT4/1O7whE+XEAkmSNcfaDW0RzTvpDA4K53EPJHL7V4eIZtkpmDQ9M1CX1FCTnK2dAd89789uxa3zB2jr/fLUO8dwLrmbqgXWz6Qv988hfB5WaweqF9cOiHsQ5PIoSjTyP6yLJAKlwWHhUNKYXBWOon1UvO7sxqvmQ7/fje77kY7oiOE4KcXnoy195yHv141JWL65HAke1LVSlnJY5kvhH5K4B4yK01//oSB2Hz/BagJkQogkQ/FsMoifPCTc/C7r50KABjZvxjnnFQdUdPPl3wlsV5p3izmjRDVnCmEM++ES7dhdRsOKHPjy5OG9Kk/yRb6VopWPpt3hJ9+CuAfq3BK1LLbz7HUuCNp4bEyuroEo6tLML22EqOrS/C9f68LK/QppfDLqXPZTCsxygI+doJlm8x0eNv8Nx77VF9+b/sxdSI3hED8w+WTsPVwO64/szah/Ul2imOr+/aTupaknjOTEUI/BfACNZy27A5h2+d98hMZVDK6Wg3AsYdI7cxYsbsJANDYEezXnWvEatMvcAR+s2zJucaPTnhvnOW7GjXzjvU9+rVpNfgalxY8USRb685nTx0r8kB1Sz+80I8nDz5//Lu3zklIn3j4co1WPPvZAQDAtiO5USgkHKHkQ6ii9gPK1HxJX5takzWZNtkL/tdfmYgzR1fp7UfbPfArqYvHSBVWnmlv3HJWGnqSGQihnwJ4gVrUB6E/c2QlxgwoibB37Ni0urmhYLbqdOTCSTVW4m7lniaMufcdbDrUFrSNCdDiAnvYa5hJsLlbOyH66MRG1DgQGiKAMBUk6/7qtsgxxArD5CPCvJMCePNMKBNOOJjQT9bDaCfhXTaZ3TofbPpWE+0faeattftbMHlYuWFbo5Y3yW4jWVM9i2m+dhtQ4FR/04FlbjR2qt8lHRP2n919XtJs++wXnTGyEmv3n0jKObKJ3H+KMwBeoL4RIow9HCz51dUzhyesTzy2COYd5taXr5p+OJ5aXQ9AnWvJHk0/UI7yt5edimvPGIEFEwehSRP66UhGNqifO2nBbexnOWNUVdj98oXcf4ozAF6g+uXYJcPAMjfqF12CS0/rm2tcKCJN5I7VTEq/v/y0pJw/k7Cy6bO2ZTvU8oAdHj+m/2YpPqkLlLe02UjWJNVXuBrEQ8sL8cDCiaguDdRyyNURXXEe18XlEeadFMAL/QevnJy+joSgyGUPWRhdVij+s64BhU573Pl+sotgqc/SBKytP4GmTi8Ot/WiucuHXy7Zru9jJ1lk3rEwF/JeSLk2kRuYt8it7xUvQuinAD6p1YDS1CcRi8SwyiJ0eiS09/iDIis3N7QBCBRpz3UsNX1u+Yvjnej0qAVG+EylhCB7zDtM0+eEPi/ocy1FAf8yLi9y9jmYLNsRQr+PePwyTnT7wu5zqLVXX2YTZ5kESwHR4TEKfUopvvrwagDA4DRkvEwHlrog13jN42v0ZT5nEgHJIpdN9T9vuuft6blmBmE/CyEEm+67IL2dyQCE0O8j33hMLRcXLZk4GeoKkWN/9d5A1OLrN89OaZ/ShVXGx1ABW7uOdQJQXWmzyKQfMO9w3/XiiYMBbAQAFLtyUywI445Kbv66KaLT48eWhnZMHFoWtkzee9uP48Nd6iRgQQqLfUeLM0SOfabVfnXKUAwoyw9N38pxJZIpuLq0AOB83jMdJvR58w6/XFyQm2JBmPRVcvPXTREsZfI3Z47AFaeHdqf8+rRhGHXP2wCME2aZglUJRd5UMeek6pT3KV3EmoYBUK8fO4pSmtJKUPHAflpzeoLTavphS0N71kxIR0u2mN1ShRD6faDHp05uRkqtwGtRRRloL2WTeLx5h192Z+A8RLKINJFrhd1GdM8QSjNfo2RxF+Zgv8Xfm4WX1zdg1sjc9GfP8J8lZQih3wd6NaFfFIMNtCoDqytZlVD0cLnhC+KIIs4lIglxp53o+yiUwpbh4sXP0mqYTI3FBQ5cl+AMmpmA0PON5I8KlwS6fap5pziGfDqZOPRnD79R6AdcNPPJv5n/qh9/oaZfeGfbsbDH2AjR5wKyQcCwUVyyUxpnGpn47KWDiL86IeRJQkgjIWQb11ZJCFlKCNmj/a/Q2gkh5K+EkDpCyBZCyFTumOu0/fcQQq5LztdJLb1RmncyHauJXPbdgPiSxGUrvE3/2ifXYtWeZuxr6g57TENrry5QEl37IBmw3zkfiuIAwMLJQwEAZ4/tn+aeZAbR/OpPAVhgarsLwAeU0rEAPtDWAeAiAGO1vxsBPAKoLwkA9wOYCWAGgPvZiyKbYTb9bPd2YA+/j5vIPdGjxh78ZP5JOL22Mi39SgdmZfB4FDUEPtJGBEB25NT35VHWVACYNqIC9YsuwajqxGeozUYi/uqU0o8BmFPTLQTwtLb8NICvcO3/piqfASgnhAwGcCGApZTSE5TSVgBLEfwiyTqYeSdSYfNMx+VQJR2v6R/SimssmDgoLX1KF2ahb5WL3cyP5o3JKhOYT8v/xH53QX4R76t+IKX0qLZ8DMBAbXkogEPcfg1aW6j2rKY3xzR9XuizXDxlYQq25yJm4R2pdjAAnDtugGEiN9Nh8zUue3YrK4L46LO0opRSQkjC7nRCyI1QTUMYPjw5qYQTwbF2D+7XEm5FY/NeecfcuHLpp4KAeYdz2dSWMzGYLJmYdV8pCqE/ZVg5PtfytEcj89t6fPBKCgamKeDtjpe3AACcQtPPS+J9oo9rZhto/xu19sMAhnH71WhtodqDoJT+g1I6nVI6vbo6c4OC/vD+bn05GsE4rLLIkL42kzBr+j5JwadaCob88/AwrvMvQisev3Y6COH89KM4x1cfWY2Z//sBWrt9eg77dJCJgYKC5BPvE70EAPPAuQ7A61z7tZoXzywA7ZoZ6D0AFxBCKrQJ3Au0tqyEUorXNqrvrG/NGpH1rmBsQo/ZehevPYj3dxw3bMsfjL9lJKHv1F6K0Zp3/vj+bt0baMqvl+L03y6Ls5/xU+Z2oH9JQdKKlggym4jmHULIYgDnAuhPCGmA6oWzCMB/CCE3ADgA4Bva7m8DuBhAHYAeAN8GAErpCULIrwF8ru33AKU0o+uWHWv34JK/roRCKZbefg76lwS09Nc3HdGH/Q8sPCVdXUwYTtNEbluPX9+WqwU1osUrhU8p3dJl1NQjmXf+9mFdX7vUZ1wOO+ZPGBh5R0FOElHoU0qvCrHpPIt9KYCbQ3zOkwCejKl3aeRQaw9atJTJ03+zDOt+fr4u+Bs7A2582a7lA1xErqbVmhOv5TOsKPy0ERVYf6DVsO1Lk4bgvPGq8NQngDmhf6LbZ9Cmo5kUTgWyouRcoRRB9OS3GhcG87CemXOA+IqbZzJ2m5pGgAl7IfQDeP3WeWocNoK/XTVFr0XAZH59i2q62d/cjam/XoonVu3XjznS1otMQFJo0PcR5A9C6IfALPR58w6buD1rTG5E+BFC4LTb4JMprnjsUzz28T4AwNC8KI9oxDxwY+Yds4y0mRqOd6hmnuv/tRZAwOzzyoYGfZ+zf79cX+bvp1Ty3JoD6PRIaSl+LsgMhNAPgdck9HlhwBTh3+VQoXCX3QavJGON5nrodtrwyV3z0tyr1GMWhVsPqwVyzP775rTETElo1eZD2JyPrNAgs870ERV460dnheyDX1YgJWm0de+rajYVe46VRBREj/jlQ2A2cfRwuWhYvdhYEq1lOgUOG9bsC8yt81k28xlWFe1AS4+h3WweMXvtsPtn17FOjL7nbUMR9UnDyjGwzI2vThmKmorg0dTYe9/BpX9blZD+8yzf1agvC5t+/iKEfgjM5p0Pdh7HIyv2YvmuRuxv7gKQW7Z9l8OGHUc70t2NjOOnF5wEAKjtX2RoN1tHzIU6eKVh3KBSPLW6Xl/XyxXaCJQQk7usFGMi2dccSByXTWkjBIklu/MHJBGWftZhI5AUimU7G7FsZ0BTIiS3olVz6bskksum1uCK04fjoy+a8EldoGawucYAn6PHLyuG5HWzRlUZhDgv9KOJ+E0UfLrszPAjEqQDIfRD0KplmVx551yUuZ2w2whe33QYd/53KwDAabPlhLsmI98ib6OFQK2BO0PLNDpuUCl2HesMiqTlZXenRzJo+mZTisTVqE1lrp5uLZ8SgKTNGQgyH/Gkh2DX0U5UFrswqMyN4gIH3E47rjh9OIb0U/OlSEpuPTR8SP7XptZg2e1z0tibzIG914dXFaF+0SWoqSiy3I8373T0+k1CX33MSt2qjjVjpJpV3E5ISn33DUI/Q2IGBKlHCP0QyApFRZEzSJu/bb5q4821Z4Y371SVuDBmQGkae5M5mAuld3hU75x/fGuaoZ0X3h0eP/76wR59nQn9GbWVWHvvebhsSg0A1bxjFvr8y2P13uYEfIMAXd6AeUfEYuQvQuiHQKHWASynDOmXht4kn355lkI5FOaXvPkWYK685uR5vOxu6fahXvP2+fd3ZuCEFtm9fHcjBpQGMmvaLDT9Xs7ufvU/18T3JULAa/olWZ4OXBA/QuiHQKHU0sNhwpAy/GjeGFyfYwWk//D1Sfqy2RMlnwj6xc1CXxPK5gyVvG3+pmfW68tnjemPusYubR/jZznsJKhIy/YjyfOgYkV/AGBSTXnSziPIbMTrPgSyEjqvzu0XnJzi3iSfimIXvn/uaDyyYm9WlPxLFWbzjl5q0DTxzbtestHAVTOGwWYjKC6wdu21EQLz1BCf7G5UdXHc/W5o7cGHuxoxfUQlJgwpA6AWxpk9pgp3XDgOk4aVx/3ZguxGaPohoJQi3xJMVhaJVLtmzO99FpvhdpqEvsWLkpluQr1E7bbgcow9nDY+ZVj8ZaQfXrEX972+HT95abPe1u2VUFLgEAI/zxGafghCmXfyAaHoBzDfAf+4dhre3no0KC+RleslmytlMR+PftM4+cu8dyil+qiy28vMR7Y+eYgxn/zdxzqgKBQ2G0G3V8760p6CvpNnumz0yDQ30ibHQp593agw3wM1FUW4cc7ooHYroc/mRn44bywqipw4Y3SVYbtDL14TEO5M0y8vckKS43/9slGGQgMeR12api/Ib4TQDwGlFCI9iSDaW8BKKWemmxkjK7HxvguCPKRGV5cAAPYc79LberUcT6Vup+FlECu8VxBLAtfjk4SmLxBCPxR5bd4R9h2daG+BsQNLgtq6PJLFngFYgZVObj+/JqwLnfY+Rc3yQr/T44dXkuGXaU4lCRTEh3jth0BW8lfo5zPDK4twzczheG7NQQDB3juh+PF5Y3HWmP6YOaoKJ7p9mPrrpbh8Wk3YY1h6Bt5275fVqlZOe9/y8vDH+iQFfs1UJIqhC4SmHwKFAiLleP5hsxH89rJTAw1RvvcddhtmjlJt9pXFLtQvugQXnTo47DEsUpePjpVkBU67DQ67rU9Rs7JJ6LNRg6iYJRBiLQQ0n807wn9HJ5m3ABP6fEZOv0zhsBEUuezo8oY3D4WDF/peWdE1f4eYqMp7hNAPQT6ad5hHirDpB0jmHeByqJ9+07PrcUCrreuXFbgcNoyoLMKB5p64o6NlheqxBF6/or8EHGL4mveIOyAEqnknz4R+ujuQgSTTbZcXwD/Vgqj8mnlnZP9idHolNHV5Qx0eFklRUORSp+x8sqKbikRtXIEQ+iFQzTvp7oUg3STzFnByqRyOdXgAAJJM4bATDK9SUzgfOtEb12criuoBBKg2fVmYdwQawnsnBHIe2vTz7OtGRTLvAb64SluPH50eP5btPA6Xw47K4gKt3RfXZ0uKgkJXQOgz7x0xkSsQmn4IFCV/64jmc5ZNM0mdyOXMO50eCX9fXocOj4TmLi8qitRArlYuAVs4JFnBJ3WB/PuyQlGmFW3p9kq6pu/Mt4RSgiDEHRACRZh3BEmmyJR9k5VgrChyYkCpG3YbQT1XzDwcz689iGseX4N3tx0FoI5Uy7UEer99e6du0xeavkAI/RDkY0TuJacNxrDKQlw/e2S6u5IxJPMWMAdK1Td3o7aqCGvuOR+FLjuGlhfi4ImeqD6ruUs1A2073AFJVrDtcAc+3NWobw947+TXPS0IRtj0Q6DQ/NOKBpS6sfKOeenuRkYRbURuIthwsA1AIFe/0x59DV1mkrPbiCGtw4iqIhxo6cHmBvWzHcK8k/fk/B3Q3uOPK8hFoVRMbArSeg84bNGnV2bFXRRK4ZHUpG33f2kC5p48AABw3+vbAQA9fQj4EuQGOS/0Jz3wPub8fnnE/ZbtOI6vPvwJrnn8MzR3edUc5ELq5z2pvgPuXDBOX3bYSdTplTs07f5vH9bpmTr7FTr1oi+MwaY6AIL8I+eFPgC9MHU4XtnYgA0H2/BJXQuW72rMS/OOIJhk11S4xJSfp4Sb3HXYok+61ukJePnM++NHANQqXwVcLMDvv3YaJouqWXlPXgj9aGho7cUZWsKs9QdacfBED4pEGtq8J9mv/YeumYqHrp6qr/O1dx12a/POtsPtuPuVrTh0IpCmodMijbPbaTNo+v2KnEH7CPKPnBb6SgypaT1+WS9y8cLnhwAAYwcE50gX5BepsPDxgp5fttuszTuL1x7E4rUHcfbvl+OiB1cCMGr6jAKH3RAA5hTRuALkuNDv0eqERoPHrwQVu542ojLRXRJkGakomckLet6NM1RO/TYuYGvXsU4c7/Cg0yNhYFkBFn31VF15cTttBnOOSLYmAHJc6HfH4KnglWTDA/fRz87FqTX9ktEtgcCAi3Oj5JftNpul0G82JWFr7/Wj0yNhzthqXDljuF6vt8Bhx/TaSoyqLgYgfPQFKkLoa3glBQWcpj+iqjgZXRIIgghl3nHaCGQLm36LyTGh1yej0+NHqVuz2WvvCTZyHdzPDUCNPREIclzoR2/e8foVfdJr3KDSZHVJIAiiIEabfotJ0+/2Suj2ySjVcu2wI9jIlbkeyyKnkgBpiMglhCwA8CAAO4DHKaWLknWuaIKyLnv4E0wYXKaZd2z44jcXiZw7gpRSYLDpc5q+XTXvePwyHlpeh5H9i3HZlKFo6zVO2rKc+0zoM/MOU2KY67HVqEGQf6RU6BNC7AAeAjAfQAOAzwkhSyilOxJ9rjc2H8EPF28Mu0+PT8LGg23YqIW/F7kcBk1LIEgFfObLYE1fwYaDrfjbh3UAgLknDwClwB0LTkavT8bfPqzDj1/YBAAo08w7uk1fM+/YmaYvZL4AqTfvzABQRyndRyn1AXgBwMJknOjBD/ZE3Od4h3GYXFks/JgFqYcvgM5r+iw4y+sPbN/b1AUAKHU78Y3pwwyfU6Z57czS4k3YZ902/ySMqCrCjJHCG02QevPOUACHuPUGADMTfZLWbh/qGrv0dZcpyZRfVvDp3hZ8tq8l0acWCGJmZP+A0wDvQcbSMHilwNwUiy4vdtkxoKxAb79qxjDMHVcNAHjo6qk4eKJH/6yJQ/vho5/NTep3EGQPGZdlkxByI4AbAWD48OFxfcahVmM62gKT//2bW47gthc36+tfnTIUWw+3Y+HkoXGdTyDoCw67DQNKC9DY6TWUM2Qum14poOkzoV/kchheED+7cJy+XlzgwPjBZSnqvSDbSLXQPwyAH5PWaG06lNJ/APgHAEyfPj0ud4MhpqRSxS7j19x9rMuwvuhrpwlbviCtDChThT6fylkNzlIM5h3mrllsKsBiXhcIQpFqof85gLGEkJFQhf2VAK5O9Emqil0YWFaAb84cgSPtHizbeVzfpigUj360V1//nzmjhMAXBDFjZCXW7j+RsvM9fu3peH/HMQzSfOoBdSJXDmXeKTA+uuaCLAJBKFIq9CmlEiHkFgDvQXXZfJJSuj3R5yGEYM095wMAfvHaNsgKRXOXF+9tP6YnVWMM5h4ygYDx7+/MQHtvdPVpE8Ggfm5ce0atoY25bFqZd9jotbaqCAeirK4lEABpsOlTSt8G8Haqzsfc3n7z5g68tukIbj1/LACgstiFE90+3eNBIOBxO+1BuehTjd2mmXcsbfpq3967bQ5EzJUgFnLeruGwqSXnWJEJVpKuuqQgzFECQfpx6i6bAfNOS7fqZszMOwWO9L+cBNlFzgt9u5apsMenCv2Pv2gCAEwYono3iHwkgkzFbrOBUqDHFxD6LLakzJ1xjneCLCHn7xym6ff6jeGIv7h0AsYMKMHCyUPS1DOBIDzMfbObE/pNnV5tW87ra4IkkfN3DvN1ppzh8/b5J6Gy2IWb544xhMALBJkES4Xc7ZWEZi9IGDkv8diDw2crHCMqYgmyAJYorccnBdImA/j71VPS1SVBDpDzQp89OHx+E1FMQpANsFFol1cyVHWbVFOeph4JcoG8FPpOEYwlyALYvdvpkVDoCnjoFIj7V9AHcv7uYVr9kXaP3mZOwCYQZCKskHlrt09PmwxARJAL+kTO3z1MW/JxAS4pqHUtEPQZu1bI/ESPUeiLlAuCvpDzQt/Kfs9HOAoEmQrT9D1+Ra+KBQhNX9A3ct4PjGlLPF6/EPqCzMfOKSyFLjvevfVsrNjdZGgXCGIl51UGPj/5/8wZBQAYVlkYaneBIGNwcAqLy27DuEFluOmc0WnskSAXyHlN/5yTqvXl2+afhOvOrA3Kty8QZCK8aVKYdASJIueF/sAyN9becx62NLTD7bQLgS/IGvhRqhD6gkSR80IfAAaUuXH+BJE3X5BdGMw7QugLEoS4kwSCDMVQL1f4GQsShBD6AkGGwtv061tEdSxBYhBCXyDIUPjiKDUVYi5KkBjywqYvEGQj4weX4Za5YzB2YAkuOXVwursjyBGE0BcIMhS7jeCnF56c7m4Icgxh3hEIBII8Qgh9gUAgyCOE0BcIBII8Qgh9gUAgyCOE0BcIBII8Qgh9gUAgyCOE0BcIBII8Qgh9gUAgyCMIpTTdfQgJIaQJwIE+fER/AM0J6k6yyaa+AtnV32zqK5Bd/c2mvgL5098RlNJqqw0ZLfT7CiFkHaV0err7EQ3Z1Fcgu/qbTX0Fsqu/2dRXQPQXEOYdgUAgyCuE0BcIBII8IteF/j/S3YEYyKa+AtnV32zqK5Bd/c2mvgKiv7lt0xcIBAKBkVzX9AUCgUDAkZNCnxCygBCymxBSRwi5KwP6M4wQspwQsoMQsp0Q8mOtvZIQspQQskf7X6G1E0LIX7X+byGETE1Tv+2EkI2EkDe19ZGEkDVav14khLi09gJtvU7bXpuGvpYTQl4mhOwihOwkhJyRqdeXEHKbdh9sI4QsJoS4M+naEkKeJIQ0EkK2cW0xX0tCyHXa/nsIIdelsK//T7sPthBCXiWElHPb7tb6upsQciHXnhKZYdVfbttPCCGUENJfW0/OtaWU5tQfADuAvQBGAXAB2AxgQpr7NBjAVG25FMAXACYA+D2Au7T2uwD8Tlu+GMA7AAiAWQDWpKnftwN4HsCb2vp/AFypLT8K4Pva8g8APKotXwngxTT09WkA39WWXQDKM/H6AhgKYD+AQu6aXp9J1xbAHABTAWzj2mK6lgAqAezT/ldoyxUp6usFABza8u+4vk7Q5EEBgJGanLCnUmZY9VdrHwbgPahxSf2TeW1T+mCm4g/AGQDe49bvBnB3uvtl6uPrAOYD2A1gsNY2GMBubfkxAFdx++v7pbCPNQA+ADAPwJvajdfMPUz6ddZu1jO0ZYe2H0lhX/tpgpSY2jPu+kIV+oe0B9ahXdsLM+3aAqg1CdKYriWAqwA8xrUb9ktmX03bLgPwnLZskAXs2qZaZlj1F8DLACYBqEdA6Cfl2uaieYc9VIwGrS0j0IbnUwCsATCQUnpU23QMwEBtORO+w18A3AFA0darALRRSiWLPun91ba3a/unipEAmgD8SzNHPU4IKUYGXl9K6WEAfwBwEMBRqNdqPTL32jJivZaZcA8DwHegastAhvaVELIQwGFK6WbTpqT0NxeFfsZCCCkB8F8At1JKO/htVH1lZ4QrFSHkUgCNlNL16e5LlDigDpkfoZROAdAN1QShkynXV7OFL4T6ohoCoBjAgrR2KkYy5VpGghByLwAJwHPp7ksoCCFFAO4BcF+qzpmLQv8wVPsYo0ZrSyuEECdUgf8cpfQVrfk4IWSwtn0wgEatPd3fYTaALxNC6gG8ANXE8yCAckKIw6JPen+17f0AtKSwvw0AGiila7T1l6G+BDLx+p4PYD+ltIlS6gfwCtTrnanXlhHrtUzrPUwIuR7ApQCu0V5SCNOndPZ1NFQFYLP2vNUA2EAIGRSmX33qby4K/c8BjNW8IVxQJ7+WpLNDhBAC4AkAOymlf+I2LQHAZt6vg2rrZ+3XarP3swC0c0PrpEMpvZtSWkMprYV6/T6klF4DYDmAy0P0l32Py7X9U6YJUkqPAThECDlZazoPwA5k5vU9CGAWIaRIuy9YXzPy2nLEei3fA3ABIaRCG91coLUlHULIAqimyS9TSntM3+FKzSNqJICxANYijTKDUrqVUjqAUlqrPW8NUJ0+jiFZ1zZZkxXp/IM66/0F1Bn5ezOgP2dBHQ5vAbBJ+7sYqm32AwB7ACwDUKntTwA8pPV/K4Dpaez7uQh474yC+pDUAXgJQIHW7tbW67Tto9LQz8kA1mnX+DWoXg0ZeX0B/ArALgDbADwD1ZskY64tgMVQ5xv8UIXQDfFcS6j29Drt79sp7GsdVJs3e9Ye5fa/V+vrbgAXce0pkRlW/TVtr0dgIjcp11ZE5AoEAkEekYvmHYFAIBCEQAh9gUAgyCOE0BcIBII8Qgh9gUAgyCOE0BcIBII8Qgh9gUAgyCOE0BcIBII8Qgh9gUAgyCP+P+iNvCww/qwJAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plt.plot(money_rec_single)\n",
    "#plt.plot(money_rec_single, label=\"cash\")\n",
    "profit_single = (np.array(money_rec_single)-cash).tolist()\n",
    "plt.plot(profit_single, label=\"profit\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "LAGyxBTZB0JS"
   },
   "source": [
    "### Adding external factors\n",
    "\n",
    "Let's go back to our equally weighted portfolio \n",
    "but add oil, gold and S&P 500 as predictor\n",
    "\n",
    "We modify our strategy by using S&P 500 prediction as well\n",
    "\n",
    "So now half of stocks must be predicted up, and S&P in 1 and 5 days also needs to be up.\n",
    "\n",
    "To predict S&P, we use S&P itself, oil and gold"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "qyLAdDs0Mo1P"
   },
   "source": [
    "#### Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "W6HvmQVNB3f9",
    "outputId": "31dfadd2-1c8a-40bf-fb10-9266dc07fcdc"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
     ]
    }
   ],
   "source": [
    "from google.colab import drive\n",
    "\n",
    "drive.mount(\"/content/drive\")\n",
    "\n",
    "path = \"drive/MyDrive/brent-daily_csv.csv\"\n",
    "df_bonus = pd.read_csv(path)\n",
    "\n",
    "#and do same thing\n",
    "df_bonus.set_index(pd.to_datetime(df_bonus[\"Date\"]),inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 455
    },
    "id": "BX9zXK-KGddX",
    "outputId": "0c42a1cc-0efd-49af-bcd7-e6941278f708"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-7a843438-1b2d-42bc-a7fb-5b450d28de54\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Date</th>\n",
       "      <th>Price</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1987-05-20</th>\n",
       "      <td>1987-05-20</td>\n",
       "      <td>18.63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1987-05-21</th>\n",
       "      <td>1987-05-21</td>\n",
       "      <td>18.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1987-05-22</th>\n",
       "      <td>1987-05-22</td>\n",
       "      <td>18.55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1987-05-25</th>\n",
       "      <td>1987-05-25</td>\n",
       "      <td>18.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1987-05-26</th>\n",
       "      <td>1987-05-26</td>\n",
       "      <td>18.63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-02-15</th>\n",
       "      <td>2023-02-15</td>\n",
       "      <td>84.11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-02-16</th>\n",
       "      <td>2023-02-16</td>\n",
       "      <td>83.84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-02-17</th>\n",
       "      <td>2023-02-17</td>\n",
       "      <td>81.97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-02-20</th>\n",
       "      <td>2023-02-20</td>\n",
       "      <td>82.79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-02-21</th>\n",
       "      <td>2023-02-21</td>\n",
       "      <td>82.14</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>9077 rows × 2 columns</p>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-7a843438-1b2d-42bc-a7fb-5b450d28de54')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-7a843438-1b2d-42bc-a7fb-5b450d28de54 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-7a843438-1b2d-42bc-a7fb-5b450d28de54');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "                  Date  Price\n",
       "Date                         \n",
       "1987-05-20  1987-05-20  18.63\n",
       "1987-05-21  1987-05-21  18.45\n",
       "1987-05-22  1987-05-22  18.55\n",
       "1987-05-25  1987-05-25  18.60\n",
       "1987-05-26  1987-05-26  18.63\n",
       "...                ...    ...\n",
       "2023-02-15  2023-02-15  84.11\n",
       "2023-02-16  2023-02-16  83.84\n",
       "2023-02-17  2023-02-17  81.97\n",
       "2023-02-20  2023-02-20  82.79\n",
       "2023-02-21  2023-02-21  82.14\n",
       "\n",
       "[9077 rows x 2 columns]"
      ]
     },
     "execution_count": 337,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_bonus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "pBvEY4EpD8Ni",
    "outputId": "90b7a238-5778-4bdb-baa6-67455585ebf0"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[*********************100%***********************]  2 of 2 completed\n"
     ]
    }
   ],
   "source": [
    "yfin.pdr_override()\n",
    "# ^ GSPC is S&P 500\n",
    "gold_data = pdr.get_data_yahoo([\"GC=F\",\"^GSPC\"])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 237
    },
    "id": "bJYOL13XFN1l",
    "outputId": "8f2902bb-0ee3-4743-a355-5441a5b1f221"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-84c08090-4f35-4eb4-8bd5-a1ab67f766d8\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>GC=F</th>\n",
       "      <th>^GSPC</th>\n",
       "      <th>Price</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2000-08-30</th>\n",
       "      <td>273.899994</td>\n",
       "      <td>1502.589966</td>\n",
       "      <td>34.13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-08-31</th>\n",
       "      <td>278.299988</td>\n",
       "      <td>1517.680054</td>\n",
       "      <td>35.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-09-01</th>\n",
       "      <td>277.000000</td>\n",
       "      <td>1520.770020</td>\n",
       "      <td>35.09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-09-05</th>\n",
       "      <td>275.799988</td>\n",
       "      <td>1507.079956</td>\n",
       "      <td>35.72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-09-06</th>\n",
       "      <td>274.200012</td>\n",
       "      <td>1492.250000</td>\n",
       "      <td>36.70</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-84c08090-4f35-4eb4-8bd5-a1ab67f766d8')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-84c08090-4f35-4eb4-8bd5-a1ab67f766d8 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-84c08090-4f35-4eb4-8bd5-a1ab67f766d8');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "                  GC=F        ^GSPC  Price\n",
       "Date                                      \n",
       "2000-08-30  273.899994  1502.589966  34.13\n",
       "2000-08-31  278.299988  1517.680054  35.08\n",
       "2000-09-01  277.000000  1520.770020  35.09\n",
       "2000-09-05  275.799988  1507.079956  35.72\n",
       "2000-09-06  274.200012  1492.250000  36.70"
      ]
     },
     "execution_count": 339,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "total_data = pd.concat([gold_data[\"Adj Close\"].dropna(),df_bonus ],axis= 1).dropna()\n",
    "total_data.drop([\"Date\"],inplace=True,axis=1)\n",
    "total_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 17
    },
    "id": "GXVqIAxhL_b5",
    "outputId": "0af47f26-5dec-4607-dfb7-05029dcc7c0f"
   },
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "    async function download(id, filename, size) {\n",
       "      if (!google.colab.kernel.accessAllowed) {\n",
       "        return;\n",
       "      }\n",
       "      const div = document.createElement('div');\n",
       "      const label = document.createElement('label');\n",
       "      label.textContent = `Downloading \"${filename}\": `;\n",
       "      div.appendChild(label);\n",
       "      const progress = document.createElement('progress');\n",
       "      progress.max = size;\n",
       "      div.appendChild(progress);\n",
       "      document.body.appendChild(div);\n",
       "\n",
       "      const buffers = [];\n",
       "      let downloaded = 0;\n",
       "\n",
       "      const channel = await google.colab.kernel.comms.open(id);\n",
       "      // Send a message to notify the kernel that we're ready.\n",
       "      channel.send({})\n",
       "\n",
       "      for await (const message of channel.messages) {\n",
       "        // Send a message to notify the kernel that we're ready.\n",
       "        channel.send({})\n",
       "        if (message.buffers) {\n",
       "          for (const buffer of message.buffers) {\n",
       "            buffers.push(buffer);\n",
       "            downloaded += buffer.byteLength;\n",
       "            progress.value = downloaded;\n",
       "          }\n",
       "        }\n",
       "      }\n",
       "      const blob = new Blob(buffers, {type: 'application/binary'});\n",
       "      const a = document.createElement('a');\n",
       "      a.href = window.URL.createObjectURL(blob);\n",
       "      a.download = filename;\n",
       "      div.appendChild(a);\n",
       "      a.click();\n",
       "      div.remove();\n",
       "    }\n",
       "  "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "download(\"download_cf2bac99-fc66-4f1f-9686-95bc054bb4ff\", \"total_data.csv\", 282456)"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from google.colab import files\n",
    "\n",
    "total_data.to_csv(\"total_data.csv\")\n",
    "files.download(\"total_data.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "MVW3dQtNHtcc"
   },
   "outputs": [],
   "source": [
    "total_data.columns = [\"Gold\", \"S&P500\",\"Oil\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 237
    },
    "id": "qzL3hAhWE4rp",
    "outputId": "2a7bf8d3-8529-40db-d017-e5f97167ef36"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-8e14ec0d-3daa-46bb-a03c-8c5b9a777eb1\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>1_day</th>\n",
       "      <th>5_day</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2000-08-30</th>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-08-31</th>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-09-01</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-09-05</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-09-06</th>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-8e14ec0d-3daa-46bb-a03c-8c5b9a777eb1')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-8e14ec0d-3daa-46bb-a03c-8c5b9a777eb1 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-8e14ec0d-3daa-46bb-a03c-8c5b9a777eb1');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "            1_day  5_day\n",
       "Date                    \n",
       "2000-08-30      1    0.0\n",
       "2000-08-31      1    0.0\n",
       "2000-09-01      0    0.0\n",
       "2000-09-05      0    0.0\n",
       "2000-09-06      1    0.0"
      ]
     },
     "execution_count": 342,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "up_data_SP500 = pd.DataFrame()\n",
    "up_data_SP500[\"1_day\"] =total_data[\"S&P500\"].diff(1).shift(-1).dropna().ge(0).astype(int)\n",
    "up_data_SP500[\"5_day\"] =total_data[\"S&P500\"].diff(5).shift(-5).dropna().ge(0).astype(int)\n",
    "\n",
    "up_data_SP500.dropna(inplace= True)\n",
    "up_data_SP500.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "sJdnqsuBFUgL"
   },
   "outputs": [],
   "source": [
    "# Need to truncate the date to match test set\n",
    "total_data_up = pd.concat([up_data_SP500,total_data.pct_change()],axis=1)\n",
    "total_data_up =total_data_up.dropna().iloc[:,: (len(up_data_SP500.columns)+len(total_data.columns))].copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "7jTxo9ePPn0n",
    "outputId": "3c21a6f3-8c8d-4fa1-9691-cdd370a402ac"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timestamp('2017-08-07 00:00:00')"
      ]
     },
     "execution_count": 344,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Date of test set is\n",
    "min(min_err_list[0][1][\"real\"].index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "DPcBxLUfP3I2"
   },
   "outputs": [],
   "source": [
    "train_set_total = total_data_up[total_data_up.index <min(min_err_list[0][1][\"real\"].index) ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "SNdPnuvgP8XZ"
   },
   "outputs": [],
   "source": [
    "# Thankfully all data are from same market so their date intersection\n",
    "# is just the date starting from test and ends at another one\n",
    "# depending on which is lower\n",
    "test_set_total= pd.concat([total_data_up,min_err_list[0][1][\"real\"]],axis=1)\n",
    "test_set_total=test_set_total.dropna().iloc[:,: (len(up_data_SP500.columns)+len(total_data.columns))].copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "OBkoQY1eOLGa"
   },
   "source": [
    "#### Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 143
    },
    "id": "HZtjaXozJl6i",
    "outputId": "9266d9ca-3de3-4bc3-873e-a01fb875026c"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-f82abca5-14b6-46f4-a9a4-eaa941df30b0\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>val</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Gold</td>\n",
       "      <td>0.338083</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>S&amp;P500</td>\n",
       "      <td>0.338927</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Oil</td>\n",
       "      <td>0.322990</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-f82abca5-14b6-46f4-a9a4-eaa941df30b0')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-f82abca5-14b6-46f4-a9a4-eaa941df30b0 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-f82abca5-14b6-46f4-a9a4-eaa941df30b0');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "     name       val\n",
       "0    Gold  0.338083\n",
       "1  S&P500  0.338927\n",
       "2     Oil  0.322990"
      ]
     },
     "execution_count": 347,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "\n",
    "clf_1_day = DecisionTreeClassifier(random_state=0)\n",
    "\n",
    "clf_1_day.fit(train_set_total.drop([\"1_day\",\"5_day\"],axis=True), train_set_total[\"1_day\"])\n",
    "\n",
    "clf_5_day = DecisionTreeClassifier(random_state=0)\n",
    "\n",
    "clf_5_day.fit(train_set_total.drop([\"1_day\",\"5_day\"],axis=True),train_set_total[\"5_day\"])\n",
    "\n",
    "pd.DataFrame({\"name\":clf_1_day.feature_names_in_, \" val\":clf_1_day.feature_importances_})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Rom-URdXONW7"
   },
   "source": [
    "#### Result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "-2mWYLYDMxTq"
   },
   "outputs": [],
   "source": [
    "n_decision = math.ceil(nstocks/2.0)\n",
    "cash = 10000\n",
    "stock_pos = [0]*nstocks\n",
    "starting_cash = cash \n",
    "hold_days = 5\n",
    "days = hold_days\n",
    "#transaction cost\n",
    "transac = 0.002\n",
    "#Cash money\n",
    "# Add min_length here because needs to account for shorter time\n",
    "# Luckily we have all the matching opening date \n",
    "# So doing loop never has this date disparity\n",
    "# Think about if one market opens on Sunday but another opens on Saturday\n",
    "min_length = min(len(min_err_list[0][1][\"real\"].index),len(test_set_total))\n",
    "\n",
    "money_record = [0]* min_length\n",
    "holding = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "R-txVSjxMuEY"
   },
   "outputs": [],
   "source": [
    "idx = 0\n",
    "for i in range(0,min_length):\n",
    "  pred = np.array([0.0]*nstocks)\n",
    "  real = np.array([0.0]*nstocks)\n",
    "  SP1days = clf_1_day.predict(pd.DataFrame(test_set_total.iloc[i,2:len(test_set_total.columns)]).transpose())\n",
    "  SP5days = clf_5_day.predict(pd.DataFrame(test_set_total.iloc[i,2:len(test_set_total.columns)]).transpose())\n",
    " \n",
    "  for j in range(0,len(min_err_list)):\n",
    "    stockInfo = min_err_list[j][1]\n",
    "    pred[j] = stockInfo[\"pred\"].iloc[i,0]\n",
    "\n",
    "    real[j] = stockInfo[\"real\"][i]\n",
    "  # number where prediction up i\n",
    "  pred_up = (sum(pred>real))\n",
    "  # now account for S&P\n",
    "  if (not holding) & (pred_up >=n_decision)& (SP1days[0] > 0 )& (SP5days[0] > 0 ):\n",
    "    holding = True \n",
    "    # position array\n",
    "    stock_pos = np.array([starting_cash*(1-transac)]*nstocks)/nstocks / real\n",
    "    money_record[i] = sum(stock_pos*real)\n",
    "    #print(\"buy\")\n",
    "    #print(sum(stock_pos*real))\n",
    "    \n",
    "  elif holding:\n",
    "    # value of portfolio\n",
    "    #print(\"hold\")\n",
    "    money_record[i] = sum(stock_pos*real)\n",
    "    days = days-1\n",
    "    # if days to hold is 0, we sell\n",
    "    if days == 0 :\n",
    "      #print(\"sell and cash\")\n",
    "      holding = False\n",
    "      days = hold_days\n",
    "\n",
    "      starting_cash = sum(stock_pos*real)*(1-transac)\n",
    "      #print(starting_cash)\n",
    "      stock_pos = [0]*nstocks\n",
    "  else :\n",
    "    money_record[i] = starting_cash\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 282
    },
    "id": "oKaOyEhQOYmK",
    "outputId": "055bb10a-c8e7-44c3-dc9a-baf7b5773734"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f291e446130>"
      ]
     },
     "execution_count": 350,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plt.plot(money_record, label=\"cash\")\n",
    "profit = (np.array(money_record)-cash).tolist()\n",
    "plt.plot(profit, label=\"profit\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Suy8yp2UUcgu"
   },
   "source": [
    "### Mean Variance optimized\n",
    "\n",
    "That kind of portfolio instead of equal weight.\n",
    "\n",
    "No rebalance/ recalculate weight at any time.\n",
    "\n",
    "Adapted from https://builtin.com/data-science/portfolio-optimization-python\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "unO8LTDfWzgw"
   },
   "source": [
    "#### Finding weight"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "7BiSSAhOWcdr",
    "outputId": "3242b802-de45-48a6-ea7e-7ff4b9361d23"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timestamp('2017-08-07 00:00:00')"
      ]
     },
     "execution_count": 351,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "min_date = min(min_err_list[0][1][\"real\"].index)\n",
    "min_date"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "yXJ-GKdsVfQE"
   },
   "outputs": [],
   "source": [
    "wholeStock = pd.DataFrame(stk_data.loc[:, ('Adj Close')])\n",
    "# We only use data before test set to find weight\n",
    "wholeStock = wholeStock[wholeStock.index <min_date].dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "cFmRJNj7XLyt",
    "outputId": "79b8d9d4-6a1f-4914-e2a2-559e6f178b35"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['CSCO', 'AAPL', 'GOOGL', 'AMZN', 'MSFT']"
      ]
     },
     "execution_count": 353,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# This is the stocks we choose nstocks minimum error to trade\n",
    "trading_stock = [i[0] for i in min_err_list]\n",
    "trading_stock"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "fBu1fq4aX9s5"
   },
   "outputs": [],
   "source": [
    "from pypfopt.expected_returns import mean_historical_return\n",
    "from pypfopt.risk_models import CovarianceShrinkage\n",
    "chosen_stock = wholeStock[trading_stock]\n",
    "# Require input as price of stock dataframe\n",
    "mu = mean_historical_return(chosen_stock)\n",
    "S = CovarianceShrinkage(chosen_stock).ledoit_wolf()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "H5ON3d9cYdjC",
    "outputId": "2c40e70b-7dc8-4b71-ae28-818bbb94e8de"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'CSCO': 0.0, 'AAPL': 0.68575, 'GOOGL': 0.20075, 'AMZN': 0.1135, 'MSFT': 0.0}\n"
     ]
    }
   ],
   "source": [
    "from pypfopt.efficient_frontier import EfficientFrontier\n",
    "\n",
    "ef = EfficientFrontier(mu, S)\n",
    "weights = ef.max_sharpe()\n",
    "\n",
    "cleaned_weights = ef.clean_weights()\n",
    "print(dict(cleaned_weights))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "6XCz7QrZY88Y"
   },
   "outputs": [],
   "source": [
    "# record the weight\n",
    "port_weight= np.array(list(cleaned_weights.values()) )\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "zMPNF0MhakEO"
   },
   "source": [
    "#### Result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "28kBShg1ajc-"
   },
   "outputs": [],
   "source": [
    "n_decision = math.ceil(nstocks/2.0)\n",
    "cash = 10000\n",
    "stock_pos = [0]*nstocks\n",
    "starting_cash = cash \n",
    "hold_days = 5\n",
    "days = hold_days\n",
    "#transaction cost\n",
    "transac = 0.002\n",
    "#Cash money\n",
    "money_record = [0]* len(min_err_list[0][1][\"real\"].index)\n",
    "holding = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "9Wkad9pCYyCl"
   },
   "outputs": [],
   "source": [
    "idx = 0\n",
    "for i in range(0,len(min_err_list[0][1][\"real\"].index)):\n",
    "  pred = np.array([0.0]*nstocks)\n",
    "  real = np.array([0.0]*nstocks)\n",
    "  for j in range(0,len(min_err_list)):\n",
    "    stockInfo = min_err_list[j][1]\n",
    "    pred[j] = stockInfo[\"pred\"].iloc[i,0]\n",
    "\n",
    "    real[j] = stockInfo[\"real\"][i]\n",
    "  # number where prediction up i\n",
    "  pred_up = (sum(pred>real))\n",
    "  if (not holding) & (pred_up >=n_decision):\n",
    "    holding = True \n",
    "    # position array\n",
    "    stock_pos = (starting_cash*(1-transac))*port_weight / real\n",
    "    money_record[i] = sum(stock_pos*real)\n",
    "    #print(\"buy\")\n",
    "    #print(sum(stock_pos*real))\n",
    "    \n",
    "  elif holding:\n",
    "    # value of portfolio\n",
    "    #print(\"hold\")\n",
    "    money_record[i] = sum(stock_pos*real)\n",
    "    days = days-1\n",
    "    # if days to hold is 0, we sell\n",
    "    if days == 0 :\n",
    "      #print(\"sell and cash\")\n",
    "      holding = False\n",
    "      days = hold_days\n",
    "\n",
    "      starting_cash = sum(stock_pos*real)*(1-transac)\n",
    "      #print(starting_cash)\n",
    "      stock_pos = [0]*nstocks\n",
    "  else :\n",
    "    money_record[i] = starting_cash\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 282
    },
    "id": "XDbcB17QapKc",
    "outputId": "193e5e3e-dfc7-411b-8142-37947a07eff8"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f29054cc6a0>"
      ]
     },
     "execution_count": 359,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plt.plot(money_record, label=\"cash\")\n",
    "profit = (np.array(money_record)-cash).tolist()\n",
    "plt.plot(profit, label=\"profit\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "0qKvjEPObasW"
   },
   "source": [
    "### Mean Variance optimized Calibrate\n",
    "\n",
    "Calibrate the weight every time we buy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "iV_9JN8Zba7k"
   },
   "outputs": [],
   "source": [
    "n_decision = math.ceil(nstocks/2.0)\n",
    "cash = 10000\n",
    "stock_pos = [0]*nstocks\n",
    "starting_cash = cash \n",
    "hold_days = 5\n",
    "days = hold_days\n",
    "#transaction cost\n",
    "transac = 0.002\n",
    "#Cash money\n",
    "money_record = [0]* len(min_err_list[0][1][\"real\"].index)\n",
    "holding = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "COS26WeUbF2Y"
   },
   "outputs": [],
   "source": [
    "idx = 0\n",
    "for i in range(0,len(min_err_list[0][1][\"real\"].index)):\n",
    "  pred = np.array([0.0]*nstocks)\n",
    "  real = np.array([0.0]*nstocks)\n",
    "  minDate = min_err_list[0][1][\"real\"].index[i]\n",
    "  for j in range(0,len(min_err_list)):\n",
    "    stockInfo = min_err_list[j][1]\n",
    "    pred[j] = stockInfo[\"pred\"].iloc[i,0]\n",
    "\n",
    "    real[j] = stockInfo[\"real\"][i]\n",
    "  # number where prediction up i\n",
    "  pred_up = (sum(pred>real))\n",
    "  if (not holding) & (pred_up >=n_decision):\n",
    "    holding = True \n",
    "    # Finding weight\n",
    "    wholeStock = pd.DataFrame(stk_data.loc[:, ('Adj Close')])\n",
    "    # We only use data before test set to find weight\n",
    "    wholeStock = wholeStock[wholeStock.index <minDate].dropna()\n",
    "    # This is the stocks we choose nstocks minimum error to trade\n",
    "    trading_stock = [i[0] for i in min_err_list]\n",
    "    trading_stock\n",
    "    chosen_stock = wholeStock[trading_stock]\n",
    "    # Require input as price of stock dataframe\n",
    "    mu = mean_historical_return(chosen_stock)\n",
    "    S = CovarianceShrinkage(chosen_stock).ledoit_wolf()\n",
    "    ef = EfficientFrontier(mu, S)\n",
    "    weights = ef.max_sharpe()\n",
    "\n",
    "    cleaned_weights = ef.clean_weights()\n",
    "    port_weight= np.array(list(cleaned_weights.values()) )\n",
    "    #print(port_weight)\n",
    "    # End of finding weight\n",
    "    # position array\n",
    "    stock_pos = (starting_cash*(1-transac))*port_weight / real\n",
    "    money_record[i] = sum(stock_pos*real)\n",
    "    #print(\"buy\")\n",
    "    #print(sum(stock_pos*real))\n",
    "    \n",
    "  elif holding:\n",
    "    # value of portfolio\n",
    "    #print(\"hold\")\n",
    "    money_record[i] = sum(stock_pos*real)\n",
    "    days = days-1\n",
    "    # if days to hold is 0, we sell\n",
    "    if days == 0 :\n",
    "      #print(\"sell and cash\")\n",
    "      holding = False\n",
    "      days = hold_days\n",
    "\n",
    "      starting_cash = sum(stock_pos*real)*(1-transac)\n",
    "      #print(starting_cash)\n",
    "      stock_pos = [0]*nstocks\n",
    "  else :\n",
    "    money_record[i] = starting_cash\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 282
    },
    "id": "TvWGyyWbchQG",
    "outputId": "d7f0ae18-a9a8-4d4a-a7d8-6f309060b6c0"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f291e3784c0>"
      ]
     },
     "execution_count": 362,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plt.plot(money_record, label=\"cash\")\n",
    "profit = (np.array(money_record)-cash).tolist()\n",
    "plt.plot(profit, label=\"profit\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "rZ2mzhOUywzo"
   },
   "source": [
    "### Rebalancing/ Benchmark portfolio\n",
    "\n",
    "A portfolio where buy at start (with no prediction) and rebalances it self to maintain a certain weight after period of time (rebalance)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "KlrztnRkzRIY"
   },
   "source": [
    "#### Equally weighted, Monthly rebalanceed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "Hj1pWyTMy4E-"
   },
   "outputs": [],
   "source": [
    "n_decision = math.ceil(nstocks/2.0)\n",
    "\n",
    "stock_pos = [0]*nstocks\n",
    "starting_cash = cash \n",
    "\n",
    "days = hold_days\n",
    "\n",
    "#Cash money\n",
    "money_record = [0]* len(min_err_list[0][1][\"real\"].index)\n",
    "holding = False\n",
    "\n",
    "\n",
    "days_to_rebal = rebal_days"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "IMo0zL_dywJA"
   },
   "outputs": [],
   "source": [
    "idx = 0\n",
    "for i in range(0,len(min_err_list[0][1][\"real\"].index)):\n",
    "  pred = np.array([0.0]*nstocks)\n",
    "  real = np.array([0.0]*nstocks)\n",
    "  for j in range(0,len(min_err_list)):\n",
    "    stockInfo = min_err_list[j][1]\n",
    "    pred[j] = stockInfo[\"pred\"].iloc[i,0]\n",
    "\n",
    "    real[j] = stockInfo[\"real\"][i]\n",
    "\n",
    "  # buy everything equally on first day\n",
    "  if i == 0:\n",
    "      stock_pos = np.array([starting_cash*(1-transac)]*nstocks)/nstocks / real\n",
    "\n",
    "  days_to_rebal = days_to_rebal -1\n",
    "\n",
    "  # rebalanced day sell everything and distribute equal weight\n",
    "  if days_to_rebal == 0:\n",
    "      starting_cash = sum(stock_pos*real)*(1-transac)\n",
    "\n",
    "      stock_pos = np.array([starting_cash*(1-transac)]*nstocks)/nstocks / real\n",
    "      days_to_rebal = rebal_days\n",
    "  money_record[i] = sum(stock_pos*real)\n",
    "    \n",
    "  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 282
    },
    "id": "JVcP37t-1SjB",
    "outputId": "668ebbfe-bdc1-48a7-c219-5958c218ca1d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f291e630370>"
      ]
     },
     "execution_count": 365,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plt.plot(money_record, label=\"cash\")\n",
    "profit = (np.array(money_record)-cash).tolist()\n",
    "plt.plot(profit, label=\"profit\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "apA_IPCb3C74"
   },
   "source": [
    "#### Mean Variance Optimized, Monthly rebalanceed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "DlImvpm33DKs"
   },
   "outputs": [],
   "source": [
    "n_decision = math.ceil(nstocks/2.0)\n",
    "\n",
    "stock_pos = [0]*nstocks\n",
    "starting_cash = cash \n",
    "\n",
    "days = hold_days\n",
    "\n",
    "#Cash money\n",
    "money_record = [0]* len(min_err_list[0][1][\"real\"].index)\n",
    "holding = False\n",
    "\n",
    "\n",
    "days_to_rebal = rebal_days"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "hr0NTyS83Lbx"
   },
   "outputs": [],
   "source": [
    "idx = 0\n",
    "for i in range(0,len(min_err_list[0][1][\"real\"].index)):\n",
    "  pred = np.array([0.0]*nstocks)\n",
    "  real = np.array([0.0]*nstocks)\n",
    "  for j in range(0,len(min_err_list)):\n",
    "    stockInfo = min_err_list[j][1]\n",
    "    pred[j] = stockInfo[\"pred\"].iloc[i,0]\n",
    "\n",
    "    real[j] = stockInfo[\"real\"][i]\n",
    "\n",
    "  # buy everything according to weight on first day\n",
    "  if i == 0:\n",
    "      # Finding weight\n",
    "      wholeStock = pd.DataFrame(stk_data.loc[:, ('Adj Close')])\n",
    "      # We only use data before test set to find weight\n",
    "      wholeStock = wholeStock[wholeStock.index <minDate].dropna()\n",
    "      # This is the stocks we choose nstocks minimum error to trade\n",
    "      trading_stock = [i[0] for i in min_err_list]\n",
    "      trading_stock\n",
    "      chosen_stock = wholeStock[trading_stock]\n",
    "      # Require input as price of stock dataframe\n",
    "      mu = mean_historical_return(chosen_stock)\n",
    "      S = CovarianceShrinkage(chosen_stock).ledoit_wolf()\n",
    "      ef = EfficientFrontier(mu, S)\n",
    "      weights = ef.max_sharpe()\n",
    "\n",
    "      cleaned_weights = ef.clean_weights()\n",
    "      port_weight= np.array(list(cleaned_weights.values()) )\n",
    "      #print(port_weight)\n",
    "      # End of finding weight\n",
    "\n",
    "      # position array\n",
    "      stock_pos = (starting_cash*(1-transac))*port_weight / real\n",
    "  \n",
    "  days_to_rebal = days_to_rebal -1\n",
    "\n",
    "  # rebalanced day sell everything and distribute equal weight\n",
    "  if days_to_rebal == 0:\n",
    "      starting_cash = sum(stock_pos*real)*(1-transac)\n",
    "\n",
    "      # Finding weight\n",
    "      wholeStock = pd.DataFrame(stk_data.loc[:, ('Adj Close')])\n",
    "      # We only use data before test set to find weight\n",
    "      wholeStock = wholeStock[wholeStock.index <minDate].dropna()\n",
    "      # This is the stocks we choose nstocks minimum error to trade\n",
    "      trading_stock = [i[0] for i in min_err_list]\n",
    "      trading_stock\n",
    "      chosen_stock = wholeStock[trading_stock]\n",
    "      # Require input as price of stock dataframe\n",
    "      mu = mean_historical_return(chosen_stock)\n",
    "      S = CovarianceShrinkage(chosen_stock).ledoit_wolf()\n",
    "      ef = EfficientFrontier(mu, S)\n",
    "      weights = ef.max_sharpe()\n",
    "\n",
    "      cleaned_weights = ef.clean_weights()\n",
    "      port_weight= np.array(list(cleaned_weights.values()) )\n",
    "      #print(port_weight)\n",
    "      # End of finding weight\n",
    "\n",
    "      # position array\n",
    "      stock_pos = (starting_cash*(1-transac))*port_weight / real\n",
    "      days_to_rebal = rebal_days\n",
    "  money_record[i] = sum(stock_pos*real)\n",
    "    \n",
    "  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 282
    },
    "id": "KxSGER5t3Dn7",
    "outputId": "fe62709d-b5e5-4040-e475-44825ba097ef"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f291efbf910>"
      ]
     },
     "execution_count": 368,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plt.plot(money_record, label=\"cash\")\n",
    "profit = (np.array(money_record)-cash).tolist()\n",
    "plt.plot(profit, label=\"profit\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "K-ezYhdF-SES"
   },
   "source": [
    "References:\n",
    "\n",
    "Tatsat, H., Puri, S. & Lookabaugh, B. (2021) Machine learning and data science blueprints for finance : from building trading strategies to robo-advisors using Python. 1st edition. Sebastopol, California, O’Reilly."
   ]
  }
 ],
 "metadata": {
  "colab": {
   "provenance": []
  },
  "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}