Stock_Algorithms/ElasticNet_Regression.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "# ElasticNet Regression"
      ],
      "metadata": {}
    },
    {
      "cell_type": "markdown",
      "source": [
        "Elastic net is a regularized regression method that use L1 and L2 penalties of the lasso and ridge methods. "
      ],
      "metadata": {}
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "\n",
        "import warnings\n",
        "warnings.filterwarnings(\"ignore\")\n",
        "\n",
        "# fix_yahoo_finance is used to fetch data \n",
        "import fix_yahoo_finance as yf\n",
        "yf.pdr_override()"
      ],
      "outputs": [],
      "execution_count": 1,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# input\n",
        "symbol = 'AMD'\n",
        "start = '2014-01-01'\n",
        "end = '2018-08-27'\n",
        "\n",
        "# Read data \n",
        "dataset = yf.download(symbol,start,end)\n",
        "\n",
        "# View Columns\n",
        "dataset.head()"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[*********************100%***********************]  1 of 1 downloaded\n"
          ]
        },
        {
          "output_type": "execute_result",
          "execution_count": 2,
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Open</th>\n",
              "      <th>High</th>\n",
              "      <th>Low</th>\n",
              "      <th>Close</th>\n",
              "      <th>Adj Close</th>\n",
              "      <th>Volume</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",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2014-01-02</th>\n",
              "      <td>3.85</td>\n",
              "      <td>3.98</td>\n",
              "      <td>3.84</td>\n",
              "      <td>3.95</td>\n",
              "      <td>3.95</td>\n",
              "      <td>20548400</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2014-01-03</th>\n",
              "      <td>3.98</td>\n",
              "      <td>4.00</td>\n",
              "      <td>3.88</td>\n",
              "      <td>4.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>22887200</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2014-01-06</th>\n",
              "      <td>4.01</td>\n",
              "      <td>4.18</td>\n",
              "      <td>3.99</td>\n",
              "      <td>4.13</td>\n",
              "      <td>4.13</td>\n",
              "      <td>42398300</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2014-01-07</th>\n",
              "      <td>4.19</td>\n",
              "      <td>4.25</td>\n",
              "      <td>4.11</td>\n",
              "      <td>4.18</td>\n",
              "      <td>4.18</td>\n",
              "      <td>42932100</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2014-01-08</th>\n",
              "      <td>4.23</td>\n",
              "      <td>4.26</td>\n",
              "      <td>4.14</td>\n",
              "      <td>4.18</td>\n",
              "      <td>4.18</td>\n",
              "      <td>30678700</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "            Open  High   Low  Close  Adj Close    Volume\n",
              "Date                                                    \n",
              "2014-01-02  3.85  3.98  3.84   3.95       3.95  20548400\n",
              "2014-01-03  3.98  4.00  3.88   4.00       4.00  22887200\n",
              "2014-01-06  4.01  4.18  3.99   4.13       4.13  42398300\n",
              "2014-01-07  4.19  4.25  4.11   4.18       4.18  42932100\n",
              "2014-01-08  4.23  4.26  4.14   4.18       4.18  30678700"
            ]
          },
          "metadata": {}
        }
      ],
      "execution_count": 2,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "X = dataset[['Open', 'High', 'Low', 'Volume']]\n",
        "Y = dataset['Adj Close']"
      ],
      "outputs": [],
      "execution_count": 3,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.linear_model import ElasticNet\n",
        "from sklearn.datasets import make_regression"
      ],
      "outputs": [],
      "execution_count": 4,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "X, Y = make_regression(n_features=2, random_state=0)"
      ],
      "outputs": [],
      "execution_count": 5,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "regr = ElasticNet(random_state=0)\n",
        "regr.fit(X, Y)"
      ],
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 6,
          "data": {
            "text/plain": [
              "ElasticNet(alpha=1.0, copy_X=True, fit_intercept=True, l1_ratio=0.5,\n",
              "      max_iter=1000, normalize=False, positive=False, precompute=False,\n",
              "      random_state=0, selection='cyclic', tol=0.0001, warm_start=False)"
            ]
          },
          "metadata": {}
        }
      ],
      "execution_count": 6,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "print(regr.coef_)"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[18.83816048 64.55968825]\n"
          ]
        }
      ],
      "execution_count": 7,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "print(regr.intercept_)"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1.4512607561654027\n"
          ]
        }
      ],
      "execution_count": 8,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "print(regr.predict([[0, 0]]))"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[1.45126076]\n"
          ]
        }
      ],
      "execution_count": 9,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "regr.score(X, Y, sample_weight=None)"
      ],
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 10,
          "data": {
            "text/plain": [
              "0.8904453086976037"
            ]
          },
          "metadata": {}
        }
      ],
      "execution_count": 10,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Stock ElasticNet Regression"
      ],
      "metadata": {}
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.cross_validation import train_test_split\n",
        "X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size = 0.2, random_state = 0)"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "C:\\Users\\Tin Hang\\Anaconda3\\envs\\py35\\lib\\site-packages\\sklearn\\cross_validation.py:41: DeprecationWarning: This module was deprecated in version 0.18 in favor of the model_selection module into which all the refactored classes and functions are moved. Also note that the interface of the new CV iterators are different from that of this module. This module will be removed in 0.20.\n",
            "  \"This module will be removed in 0.20.\", DeprecationWarning)\n"
          ]
        }
      ],
      "execution_count": 11,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "enet = ElasticNet(random_state=0)\n",
        "enet = regr.fit(X_train, Y_train)"
      ],
      "outputs": [],
      "execution_count": 12,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "Y_pred = enet.predict(X_test)"
      ],
      "outputs": [],
      "execution_count": 13,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "print(X_train.shape)\n",
        "print(Y_train.shape)\n",
        "print(X_test.shape)\n",
        "print(Y_test.shape)"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(80, 2)\n",
            "(80,)\n",
            "(20, 2)\n",
            "(20,)\n"
          ]
        }
      ],
      "execution_count": 14,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "enet.score(X_train, Y_train)"
      ],
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 16,
          "data": {
            "text/plain": [
              "0.8688235608979287"
            ]
          },
          "metadata": {}
        }
      ],
      "execution_count": 16,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# score measure the proportion of correctly classified instances\n",
        "enet.score(X_test, Y_test)"
      ],
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 18,
          "data": {
            "text/plain": [
              "0.859843606021029"
            ]
          },
          "metadata": {}
        }
      ],
      "execution_count": 18,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# coefficient is a constant by which a variable is multiplied\n",
        "enet.coef_"
      ],
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 19,
          "data": {
            "text/plain": [
              "array([19.34408276, 61.11112372])"
            ]
          },
          "metadata": {}
        }
      ],
      "execution_count": 19,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "enet.intercept_"
      ],
      "outputs": [
        {
          "output_type": "execute_result",
          "execution_count": 20,
          "data": {
            "text/plain": [
              "-0.8197636133494073"
            ]
          },
          "metadata": {}
        }
      ],
      "execution_count": 20,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import r2_score\n",
        "\n",
        "y_pred_enet = enet.fit(X_train, Y_train).predict(X_test)\n",
        "r2_score_enet = r2_score(Y_test, Y_pred)\n",
        "\n",
        "print(enet)\n",
        "print(\"r^2 on test data : %f\" % r2_score_enet)"
      ],
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "ElasticNet(alpha=1.0, copy_X=True, fit_intercept=True, l1_ratio=0.5,\n",
            "      max_iter=1000, normalize=False, positive=False, precompute=False,\n",
            "      random_state=0, selection='cyclic', tol=0.0001, warm_start=False)\n",
            "r^2 on test data : 0.859844\n"
          ]
        }
      ],
      "execution_count": 26,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(enet.coef_, color='blue', linewidth=2,\n",
        "         label='Elastic net coefficients')\n",
        "plt.legend(loc='best')\n",
        "plt.title(\"Elastic Net R^2: %f\"\n",
        "          % (r2_score_enet))\n",
        "plt.show()"
      ],
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": [
              "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\n"
            ],
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {}
        }
      ],
      "execution_count": 30,
      "metadata": {
        "collapsed": false,
        "outputHidden": false,
        "inputHidden": false
      }
    }
  ],
  "metadata": {
    "kernel_info": {
      "name": "python3"
    },
    "language_info": {
      "name": "python",
      "mimetype": "text/x-python",
      "version": "3.5.5",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py"
    },
    "kernelspec": {
      "name": "python3",
      "language": "python",
      "display_name": "Python 3"
    },
    "nteract": {
      "version": "0.12.2"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 4
}