stock_price_prediction_keras.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "stock-price-prediction-keras.ipynb",
      "provenance": [],
      "collapsed_sections": [
        "mPYS78ikjcIa"
      ],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/JosephBless/DL/blob/main/stock_price_prediction_keras.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EgaNRy_Rq0wx"
      },
      "source": [
        "# Predict stock prices  with Long short-term memory (LSTM)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "S1WPMQ_LHNQU"
      },
      "source": [
        "This simple example will show you how LSTM models predict time series data. Stock market data is a great choice for this because it's quite regular and widely available via the Internet. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vurGbUjFwBFO"
      },
      "source": [
        "## Install requirements\n",
        "We install Tensorflow 2.0 with GPU support first"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QSG622MbwEOJ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 440
        },
        "outputId": "d8b8dfef-ad87-4ef1-f3e5-d935eac50d37"
      },
      "source": [
        "!pip install tensorflow-gpu==2.0.0-alpha0"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: tensorflow-gpu==2.0.0-alpha0 in /usr/local/lib/python3.6/dist-packages (2.0.0a0)\n",
            "Requirement already satisfied: absl-py>=0.7.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0-alpha0) (0.7.1)\n",
            "Requirement already satisfied: astor>=0.6.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0-alpha0) (0.7.1)\n",
            "Requirement already satisfied: google-pasta>=0.1.2 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0-alpha0) (0.1.5)\n",
            "Requirement already satisfied: gast>=0.2.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0-alpha0) (0.2.2)\n",
            "Requirement already satisfied: keras-preprocessing>=1.0.5 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0-alpha0) (1.0.9)\n",
            "Requirement already satisfied: tb-nightly<1.14.0a20190302,>=1.14.0a20190301 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0-alpha0) (1.14.0a20190301)\n",
            "Requirement already satisfied: tf-estimator-nightly<1.14.0.dev2019030116,>=1.14.0.dev2019030115 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0-alpha0) (1.14.0.dev2019030115)\n",
            "Requirement already satisfied: wheel>=0.26 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0-alpha0) (0.33.1)\n",
            "Requirement already satisfied: keras-applications>=1.0.6 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0-alpha0) (1.0.7)\n",
            "Requirement already satisfied: protobuf>=3.6.1 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0-alpha0) (3.7.1)\n",
            "Requirement already satisfied: six>=1.10.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0-alpha0) (1.11.0)\n",
            "Requirement already satisfied: termcolor>=1.1.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0-alpha0) (1.1.0)\n",
            "Requirement already satisfied: grpcio>=1.8.6 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0-alpha0) (1.15.0)\n",
            "Requirement already satisfied: numpy<2.0,>=1.14.5 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==2.0.0-alpha0) (1.16.2)\n",
            "Requirement already satisfied: markdown>=2.6.8 in /usr/local/lib/python3.6/dist-packages (from tb-nightly<1.14.0a20190302,>=1.14.0a20190301->tensorflow-gpu==2.0.0-alpha0) (3.1)\n",
            "Requirement already satisfied: werkzeug>=0.11.15 in /usr/local/lib/python3.6/dist-packages (from tb-nightly<1.14.0a20190302,>=1.14.0a20190301->tensorflow-gpu==2.0.0-alpha0) (0.15.2)\n",
            "Requirement already satisfied: h5py in /usr/local/lib/python3.6/dist-packages (from keras-applications>=1.0.6->tensorflow-gpu==2.0.0-alpha0) (2.8.0)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.6/dist-packages (from protobuf>=3.6.1->tensorflow-gpu==2.0.0-alpha0) (40.9.0)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fmR4lwm81Vbr",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 336
        },
        "outputId": "41cd2608-3e63-48e0-f141-ab985099ed86"
      },
      "source": [
        "!pip install pandas-datareader"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: pandas-datareader in /usr/local/lib/python3.6/dist-packages (0.7.0)\n",
            "Requirement already satisfied: pandas>=0.19.2 in /usr/local/lib/python3.6/dist-packages (from pandas-datareader) (0.23.4)\n",
            "Requirement already satisfied: lxml in /usr/local/lib/python3.6/dist-packages (from pandas-datareader) (4.2.6)\n",
            "Requirement already satisfied: wrapt in /usr/local/lib/python3.6/dist-packages (from pandas-datareader) (1.10.11)\n",
            "Requirement already satisfied: requests>=2.3.0 in /usr/local/lib/python3.6/dist-packages (from pandas-datareader) (2.18.4)\n",
            "Requirement already satisfied: numpy>=1.9.0 in /usr/local/lib/python3.6/dist-packages (from pandas>=0.19.2->pandas-datareader) (1.16.2)\n",
            "Requirement already satisfied: python-dateutil>=2.5.0 in /usr/local/lib/python3.6/dist-packages (from pandas>=0.19.2->pandas-datareader) (2.5.3)\n",
            "Requirement already satisfied: pytz>=2011k in /usr/local/lib/python3.6/dist-packages (from pandas>=0.19.2->pandas-datareader) (2018.9)\n",
            "Requirement already satisfied: urllib3<1.23,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from requests>=2.3.0->pandas-datareader) (1.22)\n",
            "Requirement already satisfied: idna<2.7,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests>=2.3.0->pandas-datareader) (2.6)\n",
            "Requirement already satisfied: chardet<3.1.0,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests>=2.3.0->pandas-datareader) (3.0.4)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.6/dist-packages (from requests>=2.3.0->pandas-datareader) (2019.3.9)\n",
            "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.6/dist-packages (from python-dateutil>=2.5.0->pandas>=0.19.2->pandas-datareader) (1.11.0)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hqWZAabeiLc_",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 248
        },
        "outputId": "2212c7d1-8f7e-4743-8da5-96d67e311cd6"
      },
      "source": [
        "!apt install graphviz"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Reading package lists... Done\n",
            "Building dependency tree       \n",
            "Reading state information... Done\n",
            "graphviz is already the newest version (2.40.1-2).\n",
            "The following package was automatically installed and is no longer required:\n",
            "  libnvidia-common-410\n",
            "Use 'apt autoremove' to remove it.\n",
            "0 upgraded, 0 newly installed, 0 to remove and 6 not upgraded.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "K5Wmmx6-iPUT",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 185
        },
        "outputId": "26040899-2528-483d-fff7-b27605078dd2"
      },
      "source": [
        "!pip install pydot pydot-ng"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: pydot in /usr/local/lib/python3.6/dist-packages (1.3.0)\n",
            "Requirement already satisfied: pydot-ng in /usr/local/lib/python3.6/dist-packages (2.0.0)\n",
            "Requirement already satisfied: pyparsing>=2.1.4 in /usr/local/lib/python3.6/dist-packages (from pydot) (2.4.0)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Fcl6EKdTv51W"
      },
      "source": [
        "## Introduction"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Rui0kjzEvXTV"
      },
      "source": [
        "LSTMs are very powerful in sequence prediction problems. They can store past information."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ed7_c04-vz2f"
      },
      "source": [
        "## Loading the dataset\n",
        "I use pandas-datareader to get the historical stock prices from Yahoo! finance. For this example, I get only the historical data till the end of *training_end_data*.  "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Hh-pO_QXqysJ"
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "from pandas_datareader import data"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D8-w0eEtwnaj"
      },
      "source": [
        "tickers = 'AAPL'\n",
        "\n",
        "start_date = '1980-12-01'\n",
        "end_date = '2018-12-31'\n",
        "\n",
        "stock_data = data.get_data_yahoo(tickers, start_date, end_date)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y2HpcDRd38on",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 437
        },
        "outputId": "ba7c7768-e873-4fdd-febc-afd0945cfd80"
      },
      "source": [
        "stock_data.head(10)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>High</th>\n",
              "      <th>Low</th>\n",
              "      <th>Open</th>\n",
              "      <th>Close</th>\n",
              "      <th>Volume</th>\n",
              "      <th>Adj Close</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>1980-12-12</th>\n",
              "      <td>0.515625</td>\n",
              "      <td>0.513393</td>\n",
              "      <td>0.513393</td>\n",
              "      <td>0.513393</td>\n",
              "      <td>117258400.0</td>\n",
              "      <td>0.023007</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-15</th>\n",
              "      <td>0.488839</td>\n",
              "      <td>0.486607</td>\n",
              "      <td>0.488839</td>\n",
              "      <td>0.486607</td>\n",
              "      <td>43971200.0</td>\n",
              "      <td>0.021807</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-16</th>\n",
              "      <td>0.453125</td>\n",
              "      <td>0.450893</td>\n",
              "      <td>0.453125</td>\n",
              "      <td>0.450893</td>\n",
              "      <td>26432000.0</td>\n",
              "      <td>0.020206</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-17</th>\n",
              "      <td>0.464286</td>\n",
              "      <td>0.462054</td>\n",
              "      <td>0.462054</td>\n",
              "      <td>0.462054</td>\n",
              "      <td>21610400.0</td>\n",
              "      <td>0.020706</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-18</th>\n",
              "      <td>0.477679</td>\n",
              "      <td>0.475446</td>\n",
              "      <td>0.475446</td>\n",
              "      <td>0.475446</td>\n",
              "      <td>18362400.0</td>\n",
              "      <td>0.021307</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-19</th>\n",
              "      <td>0.506696</td>\n",
              "      <td>0.504464</td>\n",
              "      <td>0.504464</td>\n",
              "      <td>0.504464</td>\n",
              "      <td>12157600.0</td>\n",
              "      <td>0.022607</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-22</th>\n",
              "      <td>0.531250</td>\n",
              "      <td>0.529018</td>\n",
              "      <td>0.529018</td>\n",
              "      <td>0.529018</td>\n",
              "      <td>9340800.0</td>\n",
              "      <td>0.023707</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-23</th>\n",
              "      <td>0.553571</td>\n",
              "      <td>0.551339</td>\n",
              "      <td>0.551339</td>\n",
              "      <td>0.551339</td>\n",
              "      <td>11737600.0</td>\n",
              "      <td>0.024708</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-24</th>\n",
              "      <td>0.582589</td>\n",
              "      <td>0.580357</td>\n",
              "      <td>0.580357</td>\n",
              "      <td>0.580357</td>\n",
              "      <td>12000800.0</td>\n",
              "      <td>0.026008</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-26</th>\n",
              "      <td>0.636161</td>\n",
              "      <td>0.633929</td>\n",
              "      <td>0.633929</td>\n",
              "      <td>0.633929</td>\n",
              "      <td>13893600.0</td>\n",
              "      <td>0.028409</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                High       Low      Open     Close       Volume  Adj Close\n",
              "Date                                                                      \n",
              "1980-12-12  0.515625  0.513393  0.513393  0.513393  117258400.0   0.023007\n",
              "1980-12-15  0.488839  0.486607  0.488839  0.486607   43971200.0   0.021807\n",
              "1980-12-16  0.453125  0.450893  0.453125  0.450893   26432000.0   0.020206\n",
              "1980-12-17  0.464286  0.462054  0.462054  0.462054   21610400.0   0.020706\n",
              "1980-12-18  0.477679  0.475446  0.475446  0.475446   18362400.0   0.021307\n",
              "1980-12-19  0.506696  0.504464  0.504464  0.504464   12157600.0   0.022607\n",
              "1980-12-22  0.531250  0.529018  0.529018  0.529018    9340800.0   0.023707\n",
              "1980-12-23  0.553571  0.551339  0.551339  0.551339   11737600.0   0.024708\n",
              "1980-12-24  0.582589  0.580357  0.580357  0.580357   12000800.0   0.026008\n",
              "1980-12-26  0.636161  0.633929  0.633929  0.633929   13893600.0   0.028409"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_0UQ1jU7MHYi",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 344
        },
        "outputId": "a02590a4-007b-4bba-d8f4-195e292a21e4"
      },
      "source": [
        "stock_data.describe()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>High</th>\n",
              "      <th>Low</th>\n",
              "      <th>Open</th>\n",
              "      <th>Close</th>\n",
              "      <th>Volume</th>\n",
              "      <th>Adj Close</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>9594.000000</td>\n",
              "      <td>9594.000000</td>\n",
              "      <td>9594.000000</td>\n",
              "      <td>9594.000000</td>\n",
              "      <td>9.594000e+03</td>\n",
              "      <td>9594.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>26.549240</td>\n",
              "      <td>26.026566</td>\n",
              "      <td>26.297032</td>\n",
              "      <td>26.292735</td>\n",
              "      <td>8.758682e+07</td>\n",
              "      <td>22.587725</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>47.280499</td>\n",
              "      <td>46.462657</td>\n",
              "      <td>46.880676</td>\n",
              "      <td>46.878276</td>\n",
              "      <td>8.676287e+07</td>\n",
              "      <td>44.664584</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>0.198661</td>\n",
              "      <td>0.196429</td>\n",
              "      <td>0.198661</td>\n",
              "      <td>0.196429</td>\n",
              "      <td>3.472000e+05</td>\n",
              "      <td>0.008803</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>1.071429</td>\n",
              "      <td>1.031295</td>\n",
              "      <td>1.049464</td>\n",
              "      <td>1.051429</td>\n",
              "      <td>3.407435e+07</td>\n",
              "      <td>0.148196</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>1.696429</td>\n",
              "      <td>1.633929</td>\n",
              "      <td>1.664286</td>\n",
              "      <td>1.665179</td>\n",
              "      <td>5.944890e+07</td>\n",
              "      <td>0.879176</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>26.317500</td>\n",
              "      <td>25.404285</td>\n",
              "      <td>25.910000</td>\n",
              "      <td>25.841429</td>\n",
              "      <td>1.091926e+08</td>\n",
              "      <td>17.231145</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>233.470001</td>\n",
              "      <td>229.779999</td>\n",
              "      <td>230.779999</td>\n",
              "      <td>232.070007</td>\n",
              "      <td>1.855410e+09</td>\n",
              "      <td>230.275482</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "              High          Low         Open        Close        Volume  \\\n",
              "count  9594.000000  9594.000000  9594.000000  9594.000000  9.594000e+03   \n",
              "mean     26.549240    26.026566    26.297032    26.292735  8.758682e+07   \n",
              "std      47.280499    46.462657    46.880676    46.878276  8.676287e+07   \n",
              "min       0.198661     0.196429     0.198661     0.196429  3.472000e+05   \n",
              "25%       1.071429     1.031295     1.049464     1.051429  3.407435e+07   \n",
              "50%       1.696429     1.633929     1.664286     1.665179  5.944890e+07   \n",
              "75%      26.317500    25.404285    25.910000    25.841429  1.091926e+08   \n",
              "max     233.470001   229.779999   230.779999   232.070007  1.855410e+09   \n",
              "\n",
              "         Adj Close  \n",
              "count  9594.000000  \n",
              "mean     22.587725  \n",
              "std      44.664584  \n",
              "min       0.008803  \n",
              "25%       0.148196  \n",
              "50%       0.879176  \n",
              "75%      17.231145  \n",
              "max     230.275482  "
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dFKhZw-YMM0w",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 76
        },
        "outputId": "da0e55e7-f722-415b-ee9e-937edd41a9c2"
      },
      "source": [
        "stock_data_len = stock_data['Close'].count()\n",
        "print(stock_data_len)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "9594\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6jzF16Qr4qOA"
      },
      "source": [
        "I'm only interested in *close* prices"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ns7Zcn2ZClZS",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 180
        },
        "outputId": "dbae83f5-7bd4-4f55-bb48-d30a71bfa0c6"
      },
      "source": [
        "close_prices = stock_data.iloc[:, 1:2].values\n",
        "print(close_prices)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[  0.51339287]\n",
            " [  0.48660713]\n",
            " [  0.45089287]\n",
            " ...\n",
            " [150.07000732]\n",
            " [154.55000305]\n",
            " [156.47999573]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KNaLuS1X5IVA"
      },
      "source": [
        "Of course, some of the weekdays might be public holidays in which case no price will be available. For this reason, we will fill the missing prices with the latest available prices"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "P6GTsn7N5Nf2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 199
        },
        "outputId": "2c721324-d984-4df5-928c-accb5959e37e"
      },
      "source": [
        "all_bussinessdays = pd.date_range(start=start_date, end=end_date, freq='B')\n",
        "print(all_bussinessdays)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "DatetimeIndex(['1980-12-01', '1980-12-02', '1980-12-03', '1980-12-04',\n",
            "               '1980-12-05', '1980-12-08', '1980-12-09', '1980-12-10',\n",
            "               '1980-12-11', '1980-12-12',\n",
            "               ...\n",
            "               '2018-12-18', '2018-12-19', '2018-12-20', '2018-12-21',\n",
            "               '2018-12-24', '2018-12-25', '2018-12-26', '2018-12-27',\n",
            "               '2018-12-28', '2018-12-31'],\n",
            "              dtype='datetime64[ns]', length=9936, freq='B')\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NDu-0h2i9MkT"
      },
      "source": [
        "close_prices = stock_data.reindex(all_bussinessdays)\n",
        "close_prices = stock_data.fillna(method='ffill')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xTieT-gC5e-p",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 437
        },
        "outputId": "a16fcc1e-d25d-4e95-c420-6379157eb4a3"
      },
      "source": [
        "close_prices.head(10)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>High</th>\n",
              "      <th>Low</th>\n",
              "      <th>Open</th>\n",
              "      <th>Close</th>\n",
              "      <th>Volume</th>\n",
              "      <th>Adj Close</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>1980-12-12</th>\n",
              "      <td>0.515625</td>\n",
              "      <td>0.513393</td>\n",
              "      <td>0.513393</td>\n",
              "      <td>0.513393</td>\n",
              "      <td>117258400.0</td>\n",
              "      <td>0.023007</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-15</th>\n",
              "      <td>0.488839</td>\n",
              "      <td>0.486607</td>\n",
              "      <td>0.488839</td>\n",
              "      <td>0.486607</td>\n",
              "      <td>43971200.0</td>\n",
              "      <td>0.021807</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-16</th>\n",
              "      <td>0.453125</td>\n",
              "      <td>0.450893</td>\n",
              "      <td>0.453125</td>\n",
              "      <td>0.450893</td>\n",
              "      <td>26432000.0</td>\n",
              "      <td>0.020206</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-17</th>\n",
              "      <td>0.464286</td>\n",
              "      <td>0.462054</td>\n",
              "      <td>0.462054</td>\n",
              "      <td>0.462054</td>\n",
              "      <td>21610400.0</td>\n",
              "      <td>0.020706</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-18</th>\n",
              "      <td>0.477679</td>\n",
              "      <td>0.475446</td>\n",
              "      <td>0.475446</td>\n",
              "      <td>0.475446</td>\n",
              "      <td>18362400.0</td>\n",
              "      <td>0.021307</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-19</th>\n",
              "      <td>0.506696</td>\n",
              "      <td>0.504464</td>\n",
              "      <td>0.504464</td>\n",
              "      <td>0.504464</td>\n",
              "      <td>12157600.0</td>\n",
              "      <td>0.022607</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-22</th>\n",
              "      <td>0.531250</td>\n",
              "      <td>0.529018</td>\n",
              "      <td>0.529018</td>\n",
              "      <td>0.529018</td>\n",
              "      <td>9340800.0</td>\n",
              "      <td>0.023707</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-23</th>\n",
              "      <td>0.553571</td>\n",
              "      <td>0.551339</td>\n",
              "      <td>0.551339</td>\n",
              "      <td>0.551339</td>\n",
              "      <td>11737600.0</td>\n",
              "      <td>0.024708</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-24</th>\n",
              "      <td>0.582589</td>\n",
              "      <td>0.580357</td>\n",
              "      <td>0.580357</td>\n",
              "      <td>0.580357</td>\n",
              "      <td>12000800.0</td>\n",
              "      <td>0.026008</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1980-12-26</th>\n",
              "      <td>0.636161</td>\n",
              "      <td>0.633929</td>\n",
              "      <td>0.633929</td>\n",
              "      <td>0.633929</td>\n",
              "      <td>13893600.0</td>\n",
              "      <td>0.028409</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                High       Low      Open     Close       Volume  Adj Close\n",
              "Date                                                                      \n",
              "1980-12-12  0.515625  0.513393  0.513393  0.513393  117258400.0   0.023007\n",
              "1980-12-15  0.488839  0.486607  0.488839  0.486607   43971200.0   0.021807\n",
              "1980-12-16  0.453125  0.450893  0.453125  0.450893   26432000.0   0.020206\n",
              "1980-12-17  0.464286  0.462054  0.462054  0.462054   21610400.0   0.020706\n",
              "1980-12-18  0.477679  0.475446  0.475446  0.475446   18362400.0   0.021307\n",
              "1980-12-19  0.506696  0.504464  0.504464  0.504464   12157600.0   0.022607\n",
              "1980-12-22  0.531250  0.529018  0.529018  0.529018    9340800.0   0.023707\n",
              "1980-12-23  0.553571  0.551339  0.551339  0.551339   11737600.0   0.024708\n",
              "1980-12-24  0.582589  0.580357  0.580357  0.580357   12000800.0   0.026008\n",
              "1980-12-26  0.636161  0.633929  0.633929  0.633929   13893600.0   0.028409"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "06xfVbrD51Id"
      },
      "source": [
        "The dataset is now complete and free of missing values. Let's have a look to the data frame summary:"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TR3_R4OF8pNo"
      },
      "source": [
        "## Feature scaling"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hFwPXHiU8rDv"
      },
      "source": [
        "training_set = close_prices.iloc[:, 1:2].values"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GvPn9l4d-NyM",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 180
        },
        "outputId": "527eeb21-84d2-4818-c971-63369fba5cfa"
      },
      "source": [
        "print(training_set)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[  0.51339287]\n",
            " [  0.48660713]\n",
            " [  0.45089287]\n",
            " ...\n",
            " [150.07000732]\n",
            " [154.55000305]\n",
            " [156.47999573]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c9TpiTPs9QtF",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 96
        },
        "outputId": "c4e0edfc-f000-44e2-8099-00dc9f60533d"
      },
      "source": [
        "from sklearn.preprocessing import MinMaxScaler\n",
        "sc = MinMaxScaler(feature_range = (0, 1))\n",
        "training_set_scaled = sc.fit_transform(training_set)\n",
        "print(training_set_scaled.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(9594, 1)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NmYp53pYHeQu"
      },
      "source": [
        "LSTMs expect the data in a specific format, usually a 3D tensor. I start by creating data with 60 days and converting it into an array using NumPy. Next, I convert the data into a 3D dimension array with feature_set samples, 60 days and one feature at each step."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HSZBYUcIDfWV"
      },
      "source": [
        "features = []\n",
        "labels = []\n",
        "for i in range(60, stock_data_len):\n",
        "    features.append(training_set_scaled[i-60:i, 0])\n",
        "    labels.append(training_set_scaled[i, 0])\n",
        "\n",
        "features = np.array(features)\n",
        "labels = np.array(labels)\n",
        "\n",
        "features = np.reshape(features, (features.shape[0], features.shape[1], 1))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gi-GwAE_LDZm",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 76
        },
        "outputId": "2a9d22a3-9661-49f6-f9d8-462b4bedc3a2"
      },
      "source": [
        "print(labels)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.00082642 0.00089448 0.00087503 ... 0.6528062  0.67231978 0.68072627]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "atE8xU7HLuI2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 920
        },
        "outputId": "c4c08eb4-e655-48d8-ff4b-10a01e6e66a1"
      },
      "source": [
        "print(features)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[[1.38060532e-03]\n",
            "  [1.26393438e-03]\n",
            "  [1.10837330e-03]\n",
            "  ...\n",
            "  [1.13754103e-03]\n",
            "  [9.81979819e-04]\n",
            "  [8.94476613e-04]]\n",
            "\n",
            " [[1.26393438e-03]\n",
            "  [1.10837330e-03]\n",
            "  [1.15698610e-03]\n",
            "  ...\n",
            "  [9.81979819e-04]\n",
            "  [8.94476613e-04]\n",
            "  [8.26418607e-04]]\n",
            "\n",
            " [[1.10837330e-03]\n",
            "  [1.15698610e-03]\n",
            "  [1.21532157e-03]\n",
            "  ...\n",
            "  [8.94476613e-04]\n",
            "  [8.26418607e-04]\n",
            "  [8.94476613e-04]]\n",
            "\n",
            " ...\n",
            "\n",
            " [[9.85059938e-01]\n",
            "  [9.86279533e-01]\n",
            "  [1.00000000e+00]\n",
            "  ...\n",
            "  [6.50889679e-01]\n",
            "  [6.37648276e-01]\n",
            "  [6.38214540e-01]]\n",
            "\n",
            " [[9.86279533e-01]\n",
            "  [1.00000000e+00]\n",
            "  [9.86715064e-01]\n",
            "  ...\n",
            "  [6.37648276e-01]\n",
            "  [6.38214540e-01]\n",
            "  [6.52806203e-01]]\n",
            "\n",
            " [[1.00000000e+00]\n",
            "  [9.86715064e-01]\n",
            "  [9.59927459e-01]\n",
            "  ...\n",
            "  [6.38214540e-01]\n",
            "  [6.52806203e-01]\n",
            "  [6.72319776e-01]]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tfK2gzVI_sTX"
      },
      "source": [
        "Feature tensor with three dimension: features[0] contains the ..., features[1] contains the last 60 days of values and features [2] contains the  ..."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VlzVripUIyCb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 76
        },
        "outputId": "953aaa65-2d3d-421d-ce8c-87e3b3d93921"
      },
      "source": [
        "print(features.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(9534, 60, 1)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-MQYJqsDGFYA"
      },
      "source": [
        "## Create the LSTM network\n",
        "Let's create a sequenced LSTM network with 50 units. Also the net includes some dropout layers with 0.2 which means that 20% of the neurons will be dropped."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "057FLdkzDk1t"
      },
      "source": [
        "import tensorflow as tf"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Tel113HiEcYE",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 165
        },
        "outputId": "9a174bee-e77c-47b6-b7d9-9f48dbf65568"
      },
      "source": [
        "model = tf.keras.models.Sequential([\n",
        "    tf.keras.layers.LSTM(units = 50, return_sequences = True, input_shape = (features.shape[1], 1)),\n",
        "    tf.keras.layers.Dropout(0.2),\n",
        "    tf.keras.layers.LSTM(units = 50, return_sequences = True),\n",
        "    tf.keras.layers.Dropout(0.2),\n",
        "    tf.keras.layers.LSTM(units = 50, return_sequences = True),\n",
        "    tf.keras.layers.Dropout(0.2),\n",
        "    tf.keras.layers.LSTM(units = 50),\n",
        "    tf.keras.layers.Dropout(0.2),\n",
        "    tf.keras.layers.Dense(units = 1)\n",
        "])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING: Logging before flag parsing goes to stderr.\n",
            "W0414 15:18:15.979501 139980101556096 tf_logging.py:161] <tensorflow.python.keras.layers.recurrent.UnifiedLSTM object at 0x7f4f34285860>: Note that this layer is not optimized for performance. Please use tf.keras.layers.CuDNNLSTM for better performance on GPU.\n",
            "W0414 15:18:16.001110 139980101556096 tf_logging.py:161] <tensorflow.python.keras.layers.recurrent.UnifiedLSTM object at 0x7f4f34285ef0>: Note that this layer is not optimized for performance. Please use tf.keras.layers.CuDNNLSTM for better performance on GPU.\n",
            "W0414 15:18:16.006853 139980101556096 tf_logging.py:161] <tensorflow.python.keras.layers.recurrent.UnifiedLSTM object at 0x7f4f342a6748>: Note that this layer is not optimized for performance. Please use tf.keras.layers.CuDNNLSTM for better performance on GPU.\n",
            "W0414 15:18:16.013553 139980101556096 tf_logging.py:161] <tensorflow.python.keras.layers.recurrent.UnifiedLSTM object at 0x7f4f342a6e48>: Note that this layer is not optimized for performance. Please use tf.keras.layers.CuDNNLSTM for better performance on GPU.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7nlzkKHiPuaR",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 532
        },
        "outputId": "6fa7775e-c4f6-479c-80d8-2dc36f3bf7fc"
      },
      "source": [
        "print(model.summary())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "unified_lstm (UnifiedLSTM)   (None, 60, 50)            10400     \n",
            "_________________________________________________________________\n",
            "dropout (Dropout)            (None, 60, 50)            0         \n",
            "_________________________________________________________________\n",
            "unified_lstm_1 (UnifiedLSTM) (None, 60, 50)            20200     \n",
            "_________________________________________________________________\n",
            "dropout_1 (Dropout)          (None, 60, 50)            0         \n",
            "_________________________________________________________________\n",
            "unified_lstm_2 (UnifiedLSTM) (None, 60, 50)            20200     \n",
            "_________________________________________________________________\n",
            "dropout_2 (Dropout)          (None, 60, 50)            0         \n",
            "_________________________________________________________________\n",
            "unified_lstm_3 (UnifiedLSTM) (None, 50)                20200     \n",
            "_________________________________________________________________\n",
            "dropout_3 (Dropout)          (None, 50)                0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 1)                 51        \n",
            "=================================================================\n",
            "Total params: 71,051\n",
            "Trainable params: 71,051\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "None\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5W-xbus6hr3B"
      },
      "source": [
        "#tf.keras.utils.plot_model(model, to_file='my_model.png')\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Dh55JJQwpK5-"
      },
      "source": [
        "# Run tensorboard with the logdir\n",
        "#import os\n",
        "#LOG_BASE_DIR = './log'\n",
        "#os.makedirs(LOG_BASE_DIR, exist_ok=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z8f8PuP0tuxv"
      },
      "source": [
        "#!ls -l log"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MtpzRCsqvjnQ"
      },
      "source": [
        "## Load the Colab TensorBoard extention and start TensorBoard inline"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AxpGbuaSr9uc"
      },
      "source": [
        "#%load_ext tensorboard.notebook\n",
        "#%tensorboard --logdir {LOG_BASE_DIR}"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KG7MP-S0oVAv"
      },
      "source": [
        "## Define a TensorBoard callback"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ju7MktT7sX1s"
      },
      "source": [
        "#import datetime\n",
        "#logdir = os.path.join(LOG_BASE_DIR, datetime.datetime.now().strftime(\"%Y%m%d-%H%M%S\"))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eUl1ilfJnYrR"
      },
      "source": [
        "#from tensorflow.keras.callbacks import TensorBoard\n",
        "\n",
        "#tbCallBack = TensorBoard(logdir,histogram_freq=1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TVtDPeEZGmqP"
      },
      "source": [
        "The model will be compiled and optimize by the adam optimizer and set the loss function as mean_squarred_error"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lj9k7R5UFrqO"
      },
      "source": [
        "model.compile(optimizer = 'adam', loss = 'mean_squared_error')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eW4rvRXENWYC"
      },
      "source": [
        "#import os\n",
        "#print(os.environ)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3pMuiSsIOAJa"
      },
      "source": [
        "#tf.test.gpu_device_name()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "k3Od5nUSVlJ6"
      },
      "source": [
        "#from tensorflow.python.client import device_lib\n",
        "#device_lib.list_local_devices()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vdatQv4CTSvn",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1020
        },
        "outputId": "09e206c3-5194-4c31-e5fb-775654f7cc33"
      },
      "source": [
        "from time import time\n",
        "start = time()\n",
        "history = model.fit(features, labels, epochs = 20, batch_size = 32, verbose = 1)\n",
        "end = time()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/20\n",
            "9534/9534 [==============================] - 19s 2ms/sample - loss: 0.0017\n",
            "Epoch 2/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 8.3861e-04\n",
            "Epoch 3/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 8.4889e-04\n",
            "Epoch 4/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 6.7823e-04\n",
            "Epoch 5/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 6.0230e-04\n",
            "Epoch 6/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 5.9530e-04\n",
            "Epoch 7/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 5.6053e-04\n",
            "Epoch 8/20\n",
            "9534/9534 [==============================] - 18s 2ms/sample - loss: 5.6923e-04\n",
            "Epoch 9/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 4.9843e-04\n",
            "Epoch 10/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 5.7040e-04\n",
            "Epoch 11/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 4.5142e-04\n",
            "Epoch 12/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 5.1318e-04\n",
            "Epoch 13/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 4.6513e-04\n",
            "Epoch 14/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 4.5369e-04\n",
            "Epoch 15/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 4.3363e-04\n",
            "Epoch 16/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 4.8583e-04\n",
            "Epoch 17/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 5.3111e-04\n",
            "Epoch 18/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 4.6322e-04\n",
            "Epoch 19/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 4.5440e-04\n",
            "Epoch 20/20\n",
            "9534/9534 [==============================] - 17s 2ms/sample - loss: 4.6059e-04\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CMjYNddpWzrs",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 76
        },
        "outputId": "b15440ab-e26f-4e02-a479-6bab89512118"
      },
      "source": [
        "print('Total training time {} seconds'.format(end - start))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Total training time 349.84911346435547 seconds\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0jX-nM6dfJSp",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 76
        },
        "outputId": "6e80dbe3-6699-4863-9931-10540f7fcce9"
      },
      "source": [
        "#  [samples, days, features]\n",
        "print(features.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(9534, 60, 1)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yztm7-PfI_zx"
      },
      "source": [
        "testing_start_date = '2019-01-01'\n",
        "testing_end_date = '2019-04-10'\n",
        "\n",
        "test_stock_data = data.get_data_yahoo(tickers, testing_start_date, testing_end_date)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VXVbTnlWd1iY",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 280
        },
        "outputId": "d17ae639-1e8a-49ea-e593-c13d0e3fdea3"
      },
      "source": [
        "test_stock_data.tail()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>High</th>\n",
              "      <th>Low</th>\n",
              "      <th>Open</th>\n",
              "      <th>Close</th>\n",
              "      <th>Volume</th>\n",
              "      <th>Adj Close</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>2019-04-04</th>\n",
              "      <td>196.369995</td>\n",
              "      <td>193.139999</td>\n",
              "      <td>194.789993</td>\n",
              "      <td>195.690002</td>\n",
              "      <td>19114300.0</td>\n",
              "      <td>195.690002</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2019-04-05</th>\n",
              "      <td>197.100006</td>\n",
              "      <td>195.929993</td>\n",
              "      <td>196.449997</td>\n",
              "      <td>197.000000</td>\n",
              "      <td>18526600.0</td>\n",
              "      <td>197.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2019-04-08</th>\n",
              "      <td>200.229996</td>\n",
              "      <td>196.339996</td>\n",
              "      <td>196.419998</td>\n",
              "      <td>200.100006</td>\n",
              "      <td>25881700.0</td>\n",
              "      <td>200.100006</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2019-04-09</th>\n",
              "      <td>202.850006</td>\n",
              "      <td>199.229996</td>\n",
              "      <td>200.320007</td>\n",
              "      <td>199.500000</td>\n",
              "      <td>35768200.0</td>\n",
              "      <td>199.500000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2019-04-10</th>\n",
              "      <td>200.740005</td>\n",
              "      <td>198.179993</td>\n",
              "      <td>198.679993</td>\n",
              "      <td>200.619995</td>\n",
              "      <td>21695300.0</td>\n",
              "      <td>200.619995</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                  High         Low        Open       Close      Volume  \\\n",
              "Date                                                                     \n",
              "2019-04-04  196.369995  193.139999  194.789993  195.690002  19114300.0   \n",
              "2019-04-05  197.100006  195.929993  196.449997  197.000000  18526600.0   \n",
              "2019-04-08  200.229996  196.339996  196.419998  200.100006  25881700.0   \n",
              "2019-04-09  202.850006  199.229996  200.320007  199.500000  35768200.0   \n",
              "2019-04-10  200.740005  198.179993  198.679993  200.619995  21695300.0   \n",
              "\n",
              "             Adj Close  \n",
              "Date                    \n",
              "2019-04-04  195.690002  \n",
              "2019-04-05  197.000000  \n",
              "2019-04-08  200.100006  \n",
              "2019-04-09  199.500000  \n",
              "2019-04-10  200.619995  "
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 49
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1iaY63cEJzTD"
      },
      "source": [
        "test_stock_data_processed = test_stock_data.iloc[:, 1:2].values\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QdBH51wlaost",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 76
        },
        "outputId": "489e52a8-d787-4a3e-dc91-51a71cc56558"
      },
      "source": [
        "print(test_stock_data_processed.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(69, 1)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "q66EIEAXb1Pw"
      },
      "source": [
        "all_stock_data = pd.concat((stock_data['Close'], test_stock_data['Close']), axis = 0)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ugpDI3hkc8Eo"
      },
      "source": [
        "inputs = all_stock_data[len(all_stock_data) - len(test_stock_data) - 60:].values\n",
        "inputs = inputs.reshape(-1,1)\n",
        "inputs = sc.transform(inputs)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EMsXM9nEdDXS"
      },
      "source": [
        "X_test = []\n",
        "for i in range(60, 129):\n",
        "    X_test.append(inputs[i-60:i, 0])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JBo5fWfwek9K"
      },
      "source": [
        "X_test = np.array(X_test)\n",
        "X_test = np.reshape(X_test, (X_test.shape[0], X_test.shape[1], 1))\n",
        "predicted_stock_price = model.predict(X_test)\n",
        "predicted_stock_price = sc.inverse_transform(predicted_stock_price)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EuEl7MRrdLUK",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "outputId": "f0bc8c11-19c2-4eb4-af11-6189fcab8721"
      },
      "source": [
        "plt.figure(figsize=(10,6))  \n",
        "plt.plot(test_stock_data_processed, color='blue', label='Actual Apple Stock Price')  \n",
        "plt.plot(predicted_stock_price , color='red', label='Predicted Apple Stock Price')  \n",
        "plt.title('Apple Stock Price Prediction')  \n",
        "plt.xlabel('Date')  \n",
        "plt.ylabel('Apple Stock Price')  \n",
        "plt.legend()  \n",
        "plt.show()  "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmcAAAGDCAYAAABuj7cYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IgIl5LKMKgU4"
      },
      "source": [
        "#inputs = inputs.reshape(-1,1)\n",
        "#inputs = sc.transform(inputs)\n",
        "\n",
        "\n",
        "test_inputs = test_stock_data_processed.reshape(-1,1)\n",
        "test_inputs = sc.transform(test_inputs)\n",
        "\n",
        "\n",
        "print(test_inputs.shape)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1o-fpVjsdPiR"
      },
      "source": [
        "test_features = []\n",
        "for i in range(60, 291):\n",
        "    test_features.append(test_inputs[i-60:i, 0])\n",
        "    \n",
        "test_features = np.array(test_features)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EFQQ4Hkhcp2b"
      },
      "source": [
        "test_features = np.reshape(test_features, (test_features.shape[0], test_features.shape[1], 1))  \n",
        "print(test_features.shape)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TkGzHUBzKuGl"
      },
      "source": [
        "predicted_stock_price = model.predict(test_features)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FZ4hDjOXgmBX"
      },
      "source": [
        "predicted_stock_price = sc.inverse_transform(predicted_stock_price)\n",
        "print(predicted_stock_price.shape)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Fz-mr1dT1Omh"
      },
      "source": [
        "print(test_stock_data_processed.shape)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Alsqt3-lgsfj"
      },
      "source": [
        "plt.figure(figsize=(10,6))  \n",
        "plt.plot(test_stock_data_processed, color='blue', label='Actual Apple Stock Price')  \n",
        "plt.plot(predicted_stock_price , color='red', label='Predicted Apple Stock Price')  \n",
        "plt.title('Apple Stock Price Prediction')  \n",
        "plt.xlabel('Date')  \n",
        "plt.ylabel('Apple Stock Price')  \n",
        "plt.legend()  \n",
        "plt.show()  "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mPYS78ikjcIa"
      },
      "source": [
        "## Download the model and the weights"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "knvmvzsZjN1r"
      },
      "source": [
        "from google.colab import files"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zlfsWuIPi48Y"
      },
      "source": [
        "model_json = model.to_json()\n",
        "with open(\"model.json\", \"w\") as json_file:\n",
        "  json_file.write(model_json)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t_Zig-9ojIQK"
      },
      "source": [
        "files.download(\"model.json\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VQBt2bNCjVGb"
      },
      "source": [
        "model.save('weights.h5')\n",
        "files.download('weights.h5')"
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}