RL_Stock_Trading_Custom_Reward.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "57acab24",
   "metadata": {
    "colab_type": "text",
    "id": "view-in-github"
   },
   "source": [
    "<a href=\"https://colab.research.google.com/github/fshi3/RL_stock_trading/blob/master/RL_Stock_Trading_Custom_Reward.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b2bb87f7",
   "metadata": {
    "id": "b2bb87f7"
   },
   "source": [
    "<h1>Reinforcement Learning Final Project - Stock Trading Algorithm</h1>\n",
    "    \n",
    "<br>\n",
    "Implementation using the \n",
    "\"anytrading\" environment in OpenAIGym\n",
    "<br>\n",
    "https://github.com/AminHP/gym-anytrading"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "3780bdab",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "3780bdab",
    "outputId": "f0e1a918-339d-41f1-aff8-ea4551885fa9",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: gym-anytrading in /Users/forestshi/opt/anaconda3/lib/python3.7/site-packages (1.2.0)\n",
      "Requirement already satisfied: matplotlib>=3.1.1 in /Users/forestshi/opt/anaconda3/lib/python3.7/site-packages (from gym-anytrading) (3.1.3)\n",
      "Requirement already satisfied: numpy>=1.16.4 in /Users/forestshi/opt/anaconda3/lib/python3.7/site-packages (from gym-anytrading) (1.21.5)\n",
      "Requirement already satisfied: pandas>=0.24.2 in /Users/forestshi/opt/anaconda3/lib/python3.7/site-packages (from gym-anytrading) (1.0.1)\n",
      "Requirement already satisfied: gym>=0.12.5 in /Users/forestshi/opt/anaconda3/lib/python3.7/site-packages (from gym-anytrading) (0.21.0)\n",
      "Requirement already satisfied: importlib-metadata>=4.8.1 in /Users/forestshi/opt/anaconda3/lib/python3.7/site-packages (from gym>=0.12.5->gym-anytrading) (4.11.3)\n",
      "Requirement already satisfied: cloudpickle>=1.2.0 in /Users/forestshi/opt/anaconda3/lib/python3.7/site-packages (from gym>=0.12.5->gym-anytrading) (2.0.0)\n",
      "Requirement already satisfied: python-dateutil>=2.1 in /Users/forestshi/opt/anaconda3/lib/python3.7/site-packages (from matplotlib>=3.1.1->gym-anytrading) (2.8.2)\n",
      "Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /Users/forestshi/opt/anaconda3/lib/python3.7/site-packages (from matplotlib>=3.1.1->gym-anytrading) (3.0.4)\n",
      "Requirement already satisfied: cycler>=0.10 in /Users/forestshi/opt/anaconda3/lib/python3.7/site-packages (from matplotlib>=3.1.1->gym-anytrading) (0.11.0)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in /Users/forestshi/opt/anaconda3/lib/python3.7/site-packages (from matplotlib>=3.1.1->gym-anytrading) (1.3.2)\n",
      "Requirement already satisfied: pytz>=2017.2 in /Users/forestshi/opt/anaconda3/lib/python3.7/site-packages (from pandas>=0.24.2->gym-anytrading) (2021.3)\n",
      "Requirement already satisfied: zipp>=0.5 in /Users/forestshi/opt/anaconda3/lib/python3.7/site-packages (from importlib-metadata>=4.8.1->gym>=0.12.5->gym-anytrading) (3.7.0)\n",
      "Requirement already satisfied: typing-extensions>=3.6.4 in /Users/forestshi/opt/anaconda3/lib/python3.7/site-packages (from importlib-metadata>=4.8.1->gym>=0.12.5->gym-anytrading) (4.1.1)\n",
      "Requirement already satisfied: six>=1.5 in /Users/forestshi/opt/anaconda3/lib/python3.7/site-packages (from python-dateutil>=2.1->matplotlib>=3.1.1->gym-anytrading) (1.16.0)\n",
      "Note: you may need to restart the kernel to use updated packages.\n"
     ]
    }
   ],
   "source": [
    "pip install gym-anytrading"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "1eeff598",
   "metadata": {
    "id": "1eeff598"
   },
   "outputs": [],
   "source": [
    "# importing the OpenAIGym Environment for stock trading\n",
    "import gym\n",
    "import gym_anytrading\n",
    "from gym_anytrading.envs import Positions, Actions, StocksEnv\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers\n",
    "from gym_anytrading.datasets import STOCKS_GOOGL\n",
    "from gym.envs.registration import register\n",
    "import random\n",
    "from copy import deepcopy\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "n1QJRRCWQHvk",
   "metadata": {
    "id": "n1QJRRCWQHvk"
   },
   "outputs": [],
   "source": [
    "class AdvancedEnv(StocksEnv):\n",
    "\n",
    "  def __init__(self, df, window_size, frame_bound):\n",
    "      super().__init__(df, window_size, frame_bound)\n",
    "      self.trade_fee_bid_percent = 0  # unit\n",
    "      self.trade_fee_ask_percent = 0  # unit\n",
    "\n",
    "  def reset(self):\n",
    "    obs = super().reset()\n",
    "    self._total_reward = 1\n",
    "    return obs\n",
    "\n",
    "  def _update_mdd():\n",
    "    pass\n",
    "\n",
    "  def update_sharpe():\n",
    "    pass\n",
    "\n",
    "  def _get_observation(self):\n",
    "      obs = self.signal_features[(self._current_tick-self.window_size):self._current_tick]\n",
    "      obs = np.array(obs)\n",
    "      prices = obs[1:, 1]\n",
    "      changes = obs[:-1, 0]\n",
    "      opt = prices/changes\n",
    "      return opt\n",
    "\n",
    "  def step(self, action):\n",
    "    self._done = False\n",
    "    self._current_tick += 1\n",
    "\n",
    "    if self._current_tick == self._end_tick:\n",
    "            self._done = True\n",
    "\n",
    "    step_reward = self._calculate_reward(action)\n",
    "    self._total_reward *= (1+step_reward)\n",
    "\n",
    "    self._update_profit(action)\n",
    "\n",
    "    trade = False\n",
    "    if ((action == Actions.Buy.value and self._position == Positions.Short) or\n",
    "            (action == Actions.Sell.value and self._position == Positions.Long)):\n",
    "            trade = True\n",
    "    if trade:\n",
    "        self._position = self._position.opposite()\n",
    "        self._last_trade_tick = self._current_tick\n",
    "\n",
    "    self._position_history.append(self._position)\n",
    "    observation = self._get_observation()\n",
    "    info = dict(\n",
    "            total_reward = self._total_reward,\n",
    "            total_profit = self._total_profit,\n",
    "            position = self._position.value\n",
    "    )\n",
    "    self._update_history(info)\n",
    "\n",
    "    return observation, step_reward, self._done, info\n",
    "    \n",
    "  def _calculate_reward(self, action):\n",
    "        \n",
    "        current_price = self.prices[self._current_tick]\n",
    "        last_price = self.prices[self._current_tick-1]\n",
    "    \n",
    "        percent_diff = (current_price-last_price)/last_price\n",
    "        \n",
    "        if self._position == Positions.Short:\n",
    "            return float(-percent_diff)\n",
    "        else:\n",
    "            return float(percent_diff)\n",
    "            \n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "AfNIwFA5QJ7i",
   "metadata": {
    "id": "AfNIwFA5QJ7i"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/forestshi/opt/anaconda3/lib/python3.7/site-packages/gym/envs/registration.py:216: UserWarning: \u001b[33mWARN: Overriding environment advanced-stocks-v0\u001b[0m\n",
      "  logger.warn(\"Overriding environment {}\".format(id))\n"
     ]
    }
   ],
   "source": [
    "register(\n",
    "    id='advanced-stocks-v0',\n",
    "    entry_point=AdvancedEnv,\n",
    "    kwargs={\n",
    "        'df': deepcopy(STOCKS_GOOGL),\n",
    "        'window_size': 11,\n",
    "        'frame_bound': (11, len(STOCKS_GOOGL))\n",
    "    }\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "943245b2",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 455
    },
    "id": "943245b2",
    "outputId": "2efab325-333a-4027-f41b-527ea1f390f6"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>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>2009-05-22</th>\n",
       "      <td>198.528534</td>\n",
       "      <td>199.524521</td>\n",
       "      <td>196.196198</td>\n",
       "      <td>196.946945</td>\n",
       "      <td>196.946945</td>\n",
       "      <td>3433700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-05-26</th>\n",
       "      <td>196.171173</td>\n",
       "      <td>202.702698</td>\n",
       "      <td>195.195190</td>\n",
       "      <td>202.382385</td>\n",
       "      <td>202.382385</td>\n",
       "      <td>6202700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-05-27</th>\n",
       "      <td>203.023026</td>\n",
       "      <td>206.136139</td>\n",
       "      <td>202.607605</td>\n",
       "      <td>202.982986</td>\n",
       "      <td>202.982986</td>\n",
       "      <td>6062500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-05-28</th>\n",
       "      <td>204.544540</td>\n",
       "      <td>206.016022</td>\n",
       "      <td>202.507507</td>\n",
       "      <td>205.405411</td>\n",
       "      <td>205.405411</td>\n",
       "      <td>5332200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-05-29</th>\n",
       "      <td>206.261261</td>\n",
       "      <td>208.823822</td>\n",
       "      <td>205.555557</td>\n",
       "      <td>208.823822</td>\n",
       "      <td>208.823822</td>\n",
       "      <td>5291100</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-08-23</th>\n",
       "      <td>1219.880005</td>\n",
       "      <td>1235.170044</td>\n",
       "      <td>1219.349976</td>\n",
       "      <td>1221.160034</td>\n",
       "      <td>1221.160034</td>\n",
       "      <td>1233300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-08-24</th>\n",
       "      <td>1226.000000</td>\n",
       "      <td>1237.400024</td>\n",
       "      <td>1221.420044</td>\n",
       "      <td>1236.750000</td>\n",
       "      <td>1236.750000</td>\n",
       "      <td>1222700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-08-27</th>\n",
       "      <td>1244.140015</td>\n",
       "      <td>1257.869995</td>\n",
       "      <td>1240.680054</td>\n",
       "      <td>1256.270020</td>\n",
       "      <td>1256.270020</td>\n",
       "      <td>1429000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-08-28</th>\n",
       "      <td>1255.900024</td>\n",
       "      <td>1256.560059</td>\n",
       "      <td>1242.969971</td>\n",
       "      <td>1245.859985</td>\n",
       "      <td>1245.859985</td>\n",
       "      <td>1366500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-08-29</th>\n",
       "      <td>1255.000000</td>\n",
       "      <td>1267.170044</td>\n",
       "      <td>1252.800049</td>\n",
       "      <td>1264.650024</td>\n",
       "      <td>1264.650024</td>\n",
       "      <td>1846300</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2335 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   Open         High          Low        Close    Adj Close  \\\n",
       "Date                                                                          \n",
       "2009-05-22   198.528534   199.524521   196.196198   196.946945   196.946945   \n",
       "2009-05-26   196.171173   202.702698   195.195190   202.382385   202.382385   \n",
       "2009-05-27   203.023026   206.136139   202.607605   202.982986   202.982986   \n",
       "2009-05-28   204.544540   206.016022   202.507507   205.405411   205.405411   \n",
       "2009-05-29   206.261261   208.823822   205.555557   208.823822   208.823822   \n",
       "...                 ...          ...          ...          ...          ...   \n",
       "2018-08-23  1219.880005  1235.170044  1219.349976  1221.160034  1221.160034   \n",
       "2018-08-24  1226.000000  1237.400024  1221.420044  1236.750000  1236.750000   \n",
       "2018-08-27  1244.140015  1257.869995  1240.680054  1256.270020  1256.270020   \n",
       "2018-08-28  1255.900024  1256.560059  1242.969971  1245.859985  1245.859985   \n",
       "2018-08-29  1255.000000  1267.170044  1252.800049  1264.650024  1264.650024   \n",
       "\n",
       "             Volume  \n",
       "Date                 \n",
       "2009-05-22  3433700  \n",
       "2009-05-26  6202700  \n",
       "2009-05-27  6062500  \n",
       "2009-05-28  5332200  \n",
       "2009-05-29  5291100  \n",
       "...             ...  \n",
       "2018-08-23  1233300  \n",
       "2018-08-24  1222700  \n",
       "2018-08-27  1429000  \n",
       "2018-08-28  1366500  \n",
       "2018-08-29  1846300  \n",
       "\n",
       "[2335 rows x 6 columns]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Google Stock dataset\n",
    "STOCKS_GOOGL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "6cf2e657",
   "metadata": {
    "id": "6cf2e657"
   },
   "outputs": [],
   "source": [
    "env = gym.make('advanced-stocks-v0')\n",
    "env_1 = gym.make('advanced-stocks-v0',\n",
    "               df = STOCKS_GOOGL,\n",
    "               window_size = 21,\n",
    "               frame_bound = (1000, 1500))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "90715533",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 294
    },
    "id": "90715533",
    "outputId": "32ad39e3-1d04-48f7-c839-b398931c829d"
   },
   "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": [
    "env_1.reset()\n",
    "env_1.render()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "28637e41",
   "metadata": {
    "id": "28637e41"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "1197aa9f",
   "metadata": {
    "id": "1197aa9f"
   },
   "source": [
    "Max_possible_profit signifies that if the market didn't have trade fees, you could have earned 12.468122338293616 units of currency by starting with 1.0. In other words, your money increased over 1200%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "b75c3469",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "b75c3469",
    "outputId": "ed500eba-48c0-4368-c16a-9abdcde6f751"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "12.468122338293616"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "env_1.max_possible_profit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "c8ed3866",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "c8ed3866",
    "outputId": "f31e8d31-2f41-4fbb-8152-e9ed97879f68"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(array([-0.01362547, -0.02127627,  0.04433937,  0.00030004,  0.00973615,\n",
      "        0.00686972, -0.00534761, -0.00949202,  0.02201092,  0.00672721,\n",
      "       -0.00502075,  0.0111892 ,  0.0194188 ,  0.01871775, -0.00501424,\n",
      "        0.01913142, -0.00246101,  0.01004036, -0.00306737,  0.01090562]), -0.0324540971036752, False, {'total_reward': 0.9675459028963248, 'total_profit': 1.0, 'position': 1})\n"
     ]
    }
   ],
   "source": [
    "action = 1\n",
    "print(env_1.step(action))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "0de3e68f",
   "metadata": {
    "id": "0de3e68f"
   },
   "outputs": [],
   "source": [
    "observation, reward, done, info = env_1.step(action)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d1826502",
   "metadata": {
    "id": "d1826502"
   },
   "source": [
    "<h2>Random and Heuristic Methods:</h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c43cb901",
   "metadata": {
    "id": "c43cb901"
   },
   "source": [
    "<h3>Method 1: Random Buy/Sell</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "6acba4f4",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 606
    },
    "id": "6acba4f4",
    "outputId": "e526ae5f-3fab-401d-d495-34908ea234ac"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.0146305  -0.01362547 -0.02127627  0.04433937  0.00030004  0.00973615\n",
      "  0.00686972 -0.00534761 -0.00949202  0.02201092  0.00672721 -0.00502075\n",
      "  0.0111892   0.0194188   0.01871775 -0.00501424  0.01913142 -0.00246101\n",
      "  0.01004036 -0.00306737]\n",
      "info:  {'total_reward': 0.8717448432097823, 'total_profit': 1.0646435401561891, 'position': 1}\n"
     ]
    },
    {
     "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": [
    "state = env_1.reset()\n",
    "print(state)\n",
    "while True:\n",
    "    action = env_1.action_space.sample()\n",
    "    state, reward, done, info = env_1.step(action)\n",
    "\n",
    "    if done:\n",
    "        print(\"info: \", info)\n",
    "        break\n",
    "\n",
    "env_1.render_all()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "309d61fd",
   "metadata": {
    "id": "309d61fd"
   },
   "source": [
    "<h3>Method 2: Buy low sell high (last ten days moving average)</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "5b15f637",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 88
    },
    "id": "5b15f637",
    "outputId": "ecd4aae1-dece-4e06-ed19-69b6dd33c00b",
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'\\nobservation = env_1.reset()\\n\\nlast_ten_days = [] # moving average\\nmoving_avg = 0\\nwhile True:\\n    if len(last_ten_days) < 10:\\n        last_ten_days.append(observation[0][0])\\n    else:\\n        last_ten_days.pop(0)\\n        last_ten_days.append(observation[0][0])\\n    moving_avg = sum(last_ten_days) / len(last_ten_days)\\n    \\n    \\n    if observation[0][0] < moving_avg:\\n        action = 1 # buy)\\n    elif observation[0][0] > moving_avg:\\n        action = 0 # sell\\n    \\n    observation, reward, done, info = env_1.step(action)\\n    #print(type(reward))\\n    if done:\\n        print(\"info: \", info)\\n        break\\n\\nenv_1.render_all()\\n'"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "observation = env_1.reset()\n",
    "\n",
    "last_ten_days = [] # moving average\n",
    "moving_avg = 0\n",
    "while True:\n",
    "    if len(last_ten_days) < 10:\n",
    "        last_ten_days.append(observation[0][0])\n",
    "    else:\n",
    "        last_ten_days.pop(0)\n",
    "        last_ten_days.append(observation[0][0])\n",
    "    moving_avg = sum(last_ten_days) / len(last_ten_days)\n",
    "    \n",
    "    \n",
    "    if observation[0][0] < moving_avg:\n",
    "        action = 1 # buy)\n",
    "    elif observation[0][0] > moving_avg:\n",
    "        action = 0 # sell\n",
    "    \n",
    "    observation, reward, done, info = env_1.step(action)\n",
    "    #print(type(reward))\n",
    "    if done:\n",
    "        print(\"info: \", info)\n",
    "        break\n",
    "\n",
    "env_1.render_all()\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f885fa41",
   "metadata": {
    "id": "f885fa41"
   },
   "source": [
    "<h3>Method 3: Buy and Hold Strategy</h3>\n",
    "<br>\n",
    "Buy the stock at the beginning and hold throughout the period and sell stock at the very end."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "e3b0a9d4",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "e3b0a9d4",
    "outputId": "7e325e2f-6d5e-4c4f-b59d-9451b37b270b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-0.0324540971036752\n",
      "-0.013123859558008622\n",
      "0.005874730504756066\n",
      "-0.0007149116429010947\n",
      "-0.001717086270154312\n",
      "-0.019350142369590116\n",
      "-0.007454317876690034\n",
      "-0.010727299657412181\n",
      "0.009103178530074186\n",
      "-0.01470603438753689\n",
      "0.002821572352398539\n",
      "0.0005282621379610445\n",
      "-0.0041206977929952255\n",
      "-0.009831351758943997\n",
      "0.000698388329614945\n",
      "0.005746194178881114\n",
      "0.01745234675540962\n",
      "0.011924099082723169\n",
      "-0.011693727288437528\n",
      "-0.0088996400568221\n",
      "0.0057570116595735315\n",
      "-0.00223490541397262\n",
      "0.012810878392395244\n",
      "0.01621440390007061\n",
      "6.661848545525388e-05\n",
      "-0.01769778124266156\n",
      "-0.004306314041663791\n",
      "-0.01264576369460985\n",
      "-0.004127401444827464\n",
      "0.00860078555537081\n",
      "0.003914586565699556\n",
      "0.003762573903328564\n",
      "0.008530502200165877\n",
      "-0.006273422303468473\n",
      "0.004669585445257233\n",
      "0.00796454416563269\n",
      "0.012982753277002622\n",
      "0.00016572479860865177\n",
      "0.0008285573210389542\n",
      "0.015728601356181635\n",
      "0.0029992776551496145\n",
      "0.0018309407392242418\n",
      "-0.005493729794528157\n",
      "-0.001152635332835137\n",
      "-0.008567840645308154\n",
      "-0.015460997867406867\n",
      "0.015726100664851007\n",
      "-0.00757660580711721\n",
      "-0.0009957702088325526\n",
      "-0.01683470619134376\n",
      "-0.002647294935583329\n",
      "-0.003478833271679618\n",
      "0.009804290589521463\n",
      "-0.0035581698968001427\n",
      "0.018552549725472836\n",
      "0.0025989290793321635\n",
      "-0.0017318270018816832\n",
      "-0.009314892583256556\n",
      "-0.006602917889832451\n",
      "0.002256724847434939\n",
      "-0.0025204958957979673\n",
      "-0.0055030839009295495\n",
      "-0.004810806863092647\n",
      "-0.012981590234357373\n",
      "-0.011669210561451067\n",
      "-0.0031989313147201778\n",
      "0.010199471684038564\n",
      "-0.00026568910040657856\n",
      "0.00451799818110028\n",
      "0.0050383785480092045\n",
      "-0.004005947536734273\n",
      "-0.004389706310843266\n",
      "-0.018744481798886554\n",
      "-0.0018819743398271155\n",
      "0.008107966507261992\n",
      "-0.009971634300880314\n",
      "0.015916843338522148\n",
      "0.013075653915817645\n",
      "0.009097885658556164\n",
      "2.2738619579215954e-05\n",
      "0.009629636402180265\n",
      "0.0006981424204956312\n",
      "0.008462084219947394\n",
      "-0.0034925460493023023\n",
      "-0.004467813880083855\n",
      "-0.0014734820204719919\n",
      "-0.0018585636243940887\n",
      "0.019421975278135457\n",
      "-0.005457648401650231\n",
      "0.005253850978185676\n",
      "-0.01839202412916518\n",
      "0.00038352133897248995\n",
      "-0.010836237014574326\n",
      "0.001071597000904181\n",
      "-0.0020269630635591815\n",
      "-0.0005476871363327083\n",
      "0.012661086993275422\n",
      "0.0011161634532419445\n",
      "-0.013401067254036057\n",
      "-0.004269002237689759\n",
      "-0.0075771898347833395\n",
      "-0.013941841211027513\n",
      "0.00256540322441961\n",
      "0.014464984141502848\n",
      "0.004319047942805303\n",
      "0.004724849936162034\n",
      "0.006734289105767833\n",
      "0.018163100292923083\n",
      "-0.010289209106308985\n",
      "0.13796283949780352\n",
      "-0.008018492881583891\n",
      "0.00368780051909403\n",
      "0.024240270014683398\n",
      "-0.005681521896298834\n",
      "-0.010092137913247375\n",
      "-0.00019699942075121642\n",
      "0.02092607810055039\n",
      "-0.005616440693421547\n",
      "0.0001552724612284747\n",
      "-0.003434874226091083\n",
      "-0.0009056131451896513\n",
      "-0.004473153653558412\n",
      "0.00120411811677944\n",
      "-0.014470794175546165\n",
      "0.008016256486398466\n",
      "-0.005354161301747729\n",
      "0.0011775010752651412\n",
      "0.020449152369851054\n",
      "0.0026731364655025814\n",
      "-0.0016131298520856614\n",
      "-0.0019446863459572155\n",
      "-0.006155871444525023\n",
      "-0.002818953571128882\n",
      "0.011503378089972826\n",
      "-0.0021081221566618463\n",
      "0.013606003887759375\n",
      "0.011932024160135004\n",
      "0.004440630032229633\n",
      "-0.0033110744649384547\n",
      "-0.004822617208749958\n",
      "-0.0011569387998296148\n",
      "0.004671212228195473\n",
      "-0.0007937961075833673\n",
      "0.011850433210862165\n",
      "0.007729950234285587\n",
      "0.006047418142855362\n",
      "-0.006794703225394434\n",
      "-0.006804170027206401\n",
      "-0.008570433487525043\n",
      "0.011491501856521555\n",
      "-0.0029078331007462046\n",
      "0.013917712047208911\n",
      "0.0013551164718626713\n",
      "0.01325699796446684\n",
      "0.013156231980476024\n",
      "-0.002923433686886091\n",
      "0.005054671661688053\n",
      "0.0008411728851451209\n",
      "-0.007993586305736648\n",
      "0.010140041926341204\n",
      "-0.006772437794390808\n",
      "-0.007294850239444936\n",
      "0.011149379090294863\n",
      "0.01927823877904569\n",
      "0.002080977192982421\n",
      "-0.00962993926339133\n",
      "-5.308425092878423e-05\n",
      "-0.006370619764350067\n",
      "0.023526656817911402\n",
      "-0.0006785986136319745\n",
      "0.006616580257020048\n",
      "-0.004921195449034557\n",
      "0.011446937714488417\n",
      "0.0011342860758335534\n",
      "-0.004223106204969986\n",
      "-0.03126451918144208\n",
      "-0.02010985406957056\n",
      "0.01977795667846363\n",
      "-0.014327542592701026\n",
      "0.025719938526178564\n",
      "0.04014478360288817\n",
      "-0.040254992568215235\n",
      "0.004173178063334935\n",
      "0.004428198919413695\n",
      "0.014660564237848653\n",
      "0.015069537613619505\n",
      "-0.003830353433096639\n",
      "0.014706712648074112\n",
      "-0.0029322578851565797\n",
      "0.011131736406618666\n",
      "0.002416910668751853\n",
      "0.006717597127507416\n",
      "-0.0070526508745117\n",
      "0.001472092934100433\n",
      "-0.00026574954941153934\n",
      "0.007243812653973662\n",
      "0.006177219466366399\n",
      "0.00013934085087520198\n",
      "-0.0007867568357235496\n",
      "-0.0029199517986449046\n",
      "-0.010660905339401205\n",
      "0.01016051148426763\n",
      "0.002757439385204264\n",
      "0.001108110475711052\n",
      "-0.003952085711204977\n",
      "-0.00265060044745756\n",
      "-0.009557913961110852\n",
      "0.006091771427802248\n",
      "-0.015108126533857744\n",
      "-0.013674640338910032\n",
      "0.016456355932925285\n",
      "0.016072391803036663\n",
      "-0.009915254055281366\n",
      "-0.001742714176521933\n",
      "-0.01179459669460261\n",
      "-0.021224976711420977\n",
      "0.0006822339989335229\n",
      "-0.023085780187710885\n",
      "-0.015627671673139377\n",
      "0.005267956164918292\n",
      "-0.005035026136940766\n",
      "0.01828605386712126\n",
      "0.0001850362579729434\n",
      "0.005952806260074378\n",
      "-0.045931758530183726\n",
      "-0.008473168271435154\n",
      "0.031222841580907046\n",
      "0.017093806082513305\n",
      "-0.03588807983482266\n",
      "-0.01633465394891193\n",
      "0.013835171566587947\n",
      "0.006419662367872435\n",
      "0.02770186197845393\n",
      "-0.036460358441126904\n",
      "-0.00730671734589498\n",
      "0.011365120523621234\n",
      "-0.014647094408799329\n",
      "-0.0057115364232938285\n",
      "-0.021218520240930558\n",
      "-0.000229394007848348\n",
      "0.025526860511945495\n",
      "-0.002703581664346814\n",
      "0.0068240053205952805\n",
      "-0.008653248192404877\n",
      "0.0027347893938110923\n",
      "-0.02383578277845751\n",
      "-0.008745253150359301\n",
      "0.004189156370656375\n",
      "0.01239981585019683\n",
      "0.022426034165299688\n",
      "0.005776024813684649\n",
      "-0.01316616554576878\n",
      "-0.00989872619761162\n",
      "-0.0015497562136165962\n",
      "0.019931912245282908\n",
      "0.0028951579362366223\n",
      "0.01722829205125108\n",
      "0.010460250817676896\n",
      "0.015032812709706035\n",
      "0.0196346350401129\n",
      "-0.007688665330323949\n",
      "0.00019280567567075271\n",
      "0.0019104493897589472\n",
      "-0.012787539041544877\n",
      "-0.017418606743625505\n",
      "-0.0013525454662216107\n",
      "0.020171162233256972\n",
      "0.0019472076427707841\n",
      "0.008303359820508642\n",
      "-0.004257691176482389\n",
      "-0.001407686111019212\n",
      "-0.014096916299559472\n",
      "0.0015191706881143336\n",
      "-0.014366000436841253\n",
      "-0.003041812486876186\n",
      "0.018233950984653254\n",
      "0.007723071395360692\n",
      "0.002708065677411401\n",
      "0.01371523993097362\n",
      "-0.0030472410577222367\n",
      "0.023387039358847964\n",
      "-0.0019797126173056695\n",
      "0.001573237287369641\n",
      "-0.0017415680590701528\n",
      "0.011664712056887025\n",
      "-0.0012002925019914895\n",
      "0.0038931376944023014\n",
      "-0.003911794249510235\n",
      "-0.02092217785696079\n",
      "0.008575312576404573\n",
      "-0.005691180549519383\n",
      "0.01139584554635648\n",
      "0.012971935035666059\n",
      "-0.0020193383027740696\n",
      "-0.004114259953847109\n",
      "-0.01659271288301031\n",
      "0.041820146873831614\n",
      "-0.011022761424107061\n",
      "0.008572296275074368\n",
      "0.002684021706201023\n",
      "-0.0036021612928099403\n",
      "-0.008175640401060618\n",
      "0.0015717010655449586\n",
      "-0.00846383731882622\n",
      "0.0025086117853297066\n",
      "-0.02668617148096804\n",
      "-0.01026660706271987\n",
      "0.015115140102446719\n",
      "-0.01568001903996359\n",
      "0.002355401759899755\n",
      "-0.004664993379606092\n",
      "0.01072035120309294\n",
      "-0.0011938990165279058\n",
      "-0.00888697271546121\n",
      "0.021743695568619388\n",
      "0.00015400643271516875\n",
      "-0.001607802893034773\n",
      "0.01540146590116288\n",
      "0.007440480699703314\n",
      "-0.0028470667761482864\n",
      "-0.005021733151251792\n",
      "0.00020255056116163652\n",
      "-0.0033246212460621293\n",
      "-0.0041485547368815415\n",
      "-0.008705697890785055\n",
      "-0.0045969005145797045\n",
      "0.003515263949877753\n",
      "0.010766570783533566\n",
      "0.0015120109278154486\n",
      "0.00614058026392389\n",
      "0.007822797118147476\n",
      "0.006440456042736031\n",
      "-0.016056436546910583\n",
      "0.002449468843074162\n",
      "-0.0038926865730439415\n",
      "-0.010505593134245412\n",
      "-0.005573617483729096\n",
      "0.01227565816633154\n",
      "0.007659819929116526\n",
      "0.006708426145030931\n",
      "0.013611940542403224\n",
      "-0.013429143834986014\n",
      "-0.010196438455089503\n",
      "0.012246676284256519\n",
      "-0.022007926496665806\n",
      "0.004528020504058147\n",
      "-0.00015313147869503436\n",
      "0.0010206954662924466\n",
      "-0.014921514595062876\n",
      "0.00215654812417794\n",
      "0.009244585721279935\n",
      "0.002609857569296398\n",
      "-0.023274103108392685\n",
      "0.016791524826679432\n",
      "-0.02215025905626231\n",
      "-0.02736461529182978\n",
      "-0.018804376812246587\n",
      "0.007232679210646968\n",
      "-0.014507321020950517\n",
      "-0.007046025078912683\n",
      "-0.02598154742175058\n",
      "0.017993449952788964\n",
      "0.010612765218333153\n",
      "0.008661176419206747\n",
      "0.020195732295801714\n",
      "-0.008579427064198952\n",
      "0.0017853542669911643\n",
      "0.016476316501088183\n",
      "-0.0008766414968451672\n",
      "0.0032590347585130497\n",
      "0.013564843251830558\n",
      "-0.007219918354728307\n",
      "0.0007449527025220156\n",
      "-0.014604991174586468\n",
      "-0.007662577404531253\n",
      "0.0002356486424056975\n",
      "0.011616057625108766\n",
      "0.005481608135772401\n",
      "-0.0054160560835811306\n",
      "-0.00324227138378874\n",
      "-0.002246423685405709\n",
      "-0.015400109816818846\n",
      "-0.0038965405779891474\n",
      "0.004940225065834969\n",
      "-0.00628655322471018\n",
      "0.003917178462609077\n",
      "0.002912610519171875\n",
      "0.003196463914534373\n",
      "-0.0027310963614914103\n",
      "0.002464785659532537\n",
      "-0.017174168988196792\n",
      "-0.0019642304324257133\n",
      "-0.003007957664986576\n",
      "0.010447597264242568\n",
      "-0.026724168870376958\n",
      "0.005018108837093067\n",
      "0.010136990942173614\n",
      "-0.015052894416805013\n",
      "0.007707762990627044\n",
      "-0.019920646668564156\n",
      "-0.010872241934531617\n",
      "-0.03427423634188037\n",
      "0.016641255687801063\n",
      "0.01613186456001111\n",
      "0.010532010129143936\n",
      "0.02357512983357592\n",
      "0.012154860315345453\n",
      "-0.0034152364305647096\n",
      "0.008548650773547078\n",
      "-0.00777445310332198\n",
      "-0.003778022012536701\n",
      "-0.008631101011990074\n",
      "-0.002091706660528646\n",
      "-0.019053849926628653\n",
      "-0.024679487269570706\n",
      "-0.0029409856227009877\n",
      "0.003484133467098419\n",
      "-0.01221124647601164\n",
      "-0.007309480333700319\n",
      "0.009536052024045436\n",
      "0.008230380826553514\n",
      "-0.0037949578529929183\n",
      "0.012797327180069373\n",
      "-0.0010186674943542568\n",
      "0.020492632386192224\n",
      "0.032494806803700725\n",
      "0.008654427887312795\n",
      "-0.009650412185985942\n",
      "-0.02893495647472386\n",
      "-0.016807707297501015\n",
      "0.0015611634973130611\n",
      "0.047386179583663514\n",
      "-0.009952518127486164\n",
      "0.002066847003378064\n",
      "-0.013500866607932524\n",
      "0.007089985117201428\n",
      "0.007643938376560467\n",
      "-0.008616123392746223\n",
      "0.020556120397280327\n",
      "-0.003998765380566317\n",
      "0.014888494423791757\n",
      "0.009431993746781438\n",
      "-0.011158217761216253\n",
      "-0.004330170009170906\n",
      "0.007002649648827777\n",
      "-0.008509513949832305\n",
      "-0.012550734866387603\n",
      "0.0068224747663551966\n",
      "0.016114346260569307\n",
      "0.021851461875879576\n",
      "0.005971905686462995\n",
      "0.022021603700286236\n",
      "0.006556220425160108\n",
      "-0.0007946941334219528\n",
      "0.005360219786067182\n",
      "-0.01467067248455471\n",
      "0.002094522516549832\n",
      "-0.024821460713664967\n",
      "-0.007430515635138715\n",
      "0.009861579262316953\n",
      "-0.01455883822638461\n",
      "0.015623896925859063\n",
      "-0.007175468079852045\n",
      "0.015333276358026659\n",
      "-0.004398032568084841\n",
      "0.002270883741559843\n",
      "0.0007433985150530486\n",
      "0.021525696636943037\n",
      "-0.018249780796992757\n",
      "-0.005925899470899361\n",
      "-0.01080481661686152\n",
      "0.006438932969719797\n",
      "-0.011476641885893748\n",
      "-0.009392503853055588\n",
      "-0.014886516859024073\n",
      "0.0048770833898396804\n",
      "0.0016728982074182262\n",
      "0.00730470599708295\n",
      "-0.0014940728803657733\n",
      "0.0009487938050145944\n",
      "0.00018236957015381475\n",
      "-0.01614899707962433\n",
      "0.0023341897297204485\n",
      "0.004583842416169893\n",
      "-0.01983373123955951\n",
      "0.02213094421539479\n",
      "-0.0029567625553300956\n",
      "0.011530262646456993\n",
      "0.015077076924759736\n",
      "0.029060291968345052\n",
      "-0.01314363622159153\n",
      "-0.003091217437038238\n",
      "-0.005280188575581441\n",
      "-0.02247990641586316\n",
      "0.004355108538910328\n",
      "0.0030482148238292466\n",
      "0.017726735622201874\n",
      "0.01465814916484845\n",
      "-0.013007327462950314\n",
      "info:  {'total_reward': 1.189290845617156, 'total_profit': 1.2060120246225392, 'position': 0}\n"
     ]
    },
    {
     "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": [
    "state = env_1.reset()\n",
    "count = 0\n",
    "while True:\n",
    "    count += 1\n",
    "    action = 1\n",
    "    if count > 495:\n",
    "        action = 0\n",
    "    state, reward, done, info = env_1.step(action)\n",
    "    print(reward)\n",
    "    if done:\n",
    "        print(\"info: \", info)\n",
    "        break\n",
    "\n",
    "env_1.render_all()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7df050ff",
   "metadata": {
    "id": "7df050ff"
   },
   "source": [
    "<h2>Reinforcement Learning Methods:</h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5c5becd4",
   "metadata": {
    "id": "5c5becd4"
   },
   "source": [
    "<h3>Method 4: Deep Q-Learning (DQN)</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "d5dfa0ae",
   "metadata": {
    "id": "d5dfa0ae"
   },
   "outputs": [],
   "source": [
    "# Parameters\n",
    "gamma = 0.99\n",
    "actsize = env_1.action_space.n\n",
    "obssize = env_1.observation_space.low.size\n",
    "window_size = 20\n",
    "\n",
    "tf.random.set_seed(69420)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "3994d96b",
   "metadata": {
    "id": "3994d96b"
   },
   "outputs": [],
   "source": [
    "def model_creator():\n",
    "    model = keras.Sequential()\n",
    "    model.add(layers.Reshape((1,window_size), input_shape=(window_size,)))\n",
    "    model.add(layers.LSTM(1, return_sequences=True))\n",
    "    model.add(layers.LSTM(1, return_sequences=False))\n",
    "    #model.add(layers.Dense(128,activation=\"relu\",  dtype='float64', input_shape=(window_size,)))\n",
    "    #model.add(layers.Dense(500,activation=\"relu\"))\n",
    "    #model.add(layers.Dense(1000,activation=\"relu\"))\n",
    "    #model.add(layers.Dense(30))\n",
    "    model.add(layers.Dense(actsize))\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "0b0d9905",
   "metadata": {
    "id": "0b0d9905"
   },
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "Input 0 of layer lstm_2 is incompatible with the layer: expected ndim=3, found ndim=2. Full shape received: [10, 42]",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-31-25225ccc69bc>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0mmodel_target\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel_creator\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      6\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 7\u001b[0;31m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbuild\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch_size\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mobssize\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      8\u001b[0m \u001b[0mmodel_target\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbuild\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch_size\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mobssize\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/tensorflow/python/keras/engine/sequential.py\u001b[0m in \u001b[0;36mbuild\u001b[0;34m(self, input_shape)\u001b[0m\n\u001b[1;32m    347\u001b[0m         \u001b[0minput_shape\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtuple\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput_shape\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    348\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_build_input_shape\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0minput_shape\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 349\u001b[0;31m         \u001b[0msuper\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mSequential\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbuild\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput_shape\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    350\u001b[0m     \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbuilt\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    351\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/tensorflow/python/keras/engine/training.py\u001b[0m in \u001b[0;36mbuild\u001b[0;34m(self, input_shape)\u001b[0m\n\u001b[1;32m    430\u001b[0m                            'method accepts an `inputs` argument.')\n\u001b[1;32m    431\u001b[0m         \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 432\u001b[0;31m           \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcall\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    433\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0merrors\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mInvalidArgumentError\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mTypeError\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    434\u001b[0m           raise ValueError('You cannot build your model by calling `build` '\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/tensorflow/python/keras/engine/sequential.py\u001b[0m in \u001b[0;36mcall\u001b[0;34m(self, inputs, training, mask)\u001b[0m\n\u001b[1;32m    384\u001b[0m         \u001b[0mkwargs\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'training'\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtraining\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    385\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 386\u001b[0;31m       \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlayer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    387\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    388\u001b[0m       \u001b[0;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnest\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mflatten\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m!=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/tensorflow/python/keras/layers/recurrent.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, inputs, initial_state, constants, **kwargs)\u001b[0m\n\u001b[1;32m    661\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    662\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0minitial_state\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0mconstants\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 663\u001b[0;31m       \u001b[0;32mreturn\u001b[0m \u001b[0msuper\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mRNN\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__call__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    664\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    665\u001b[0m     \u001b[0;31m# If any of `initial_state` or `constants` are specified and are Keras\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/tensorflow/python/keras/engine/base_layer.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m    974\u001b[0m       \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    975\u001b[0m         input_spec.assert_input_compatibility(self.input_spec, inputs,\n\u001b[0;32m--> 976\u001b[0;31m                                               self.name)\n\u001b[0m\u001b[1;32m    977\u001b[0m         \u001b[0mname_scope\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_name_scope\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;31m# Avoid autoincrementing.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    978\u001b[0m         \u001b[0mcall_fn\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_autographed_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/tensorflow/python/keras/engine/input_spec.py\u001b[0m in \u001b[0;36massert_input_compatibility\u001b[0;34m(input_spec, inputs, layer_name)\u001b[0m\n\u001b[1;32m    178\u001b[0m                          \u001b[0;34m'expected ndim='\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0mstr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mspec\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mndim\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0;34m', found ndim='\u001b[0m \u001b[0;34m+\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    179\u001b[0m                          \u001b[0mstr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mndim\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0;34m'. Full shape received: '\u001b[0m \u001b[0;34m+\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 180\u001b[0;31m                          str(x.shape.as_list()))\n\u001b[0m\u001b[1;32m    181\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mspec\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmax_ndim\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    182\u001b[0m       \u001b[0mndim\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mndims\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mValueError\u001b[0m: Input 0 of layer lstm_2 is incompatible with the layer: expected ndim=3, found ndim=2. Full shape received: [10, 42]"
     ]
    }
   ],
   "source": [
    "batch_size = 10\n",
    "# Model used for selecting actions (principal)\n",
    "model = model_creator()\n",
    "# Then create the target model. This will periodically be copied from the principal network \n",
    "model_target = model_creator()\n",
    "\n",
    "model.build((batch_size,obssize,))\n",
    "model_target.build((batch_size,obssize,))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f6cdc499",
   "metadata": {
    "id": "f6cdc499"
   },
   "source": [
    "**Training the Model:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "57730d1a",
   "metadata": {
    "id": "57730d1a"
   },
   "outputs": [],
   "source": [
    "# environment for training\n",
    "env_training = gym.make('advanced-stocks-v0', df = STOCKS_GOOGL, window_size = window_size+1, frame_bound = (window_size+1, 1000))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "d1820fde",
   "metadata": {
    "id": "d1820fde"
   },
   "outputs": [],
   "source": [
    "#tf.keras.backend.set_floatx('float64')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "7dc9132e",
   "metadata": {
    "id": "7dc9132e"
   },
   "outputs": [],
   "source": [
    "optimizer = keras.optimizers.SGD(learning_rate=0.0005)\n",
    "\n",
    "\n",
    "    # Our Experience Replay memory \n",
    "action_history = []\n",
    "state_history = []\n",
    "state_next_history = []\n",
    "rewards_history = []\n",
    "done_history = []\n",
    "running_rewards_history = []\n",
    "\n",
    "# Replay memory size\n",
    "max_memory = 100 # You can experiment with different sizes.\n",
    "\n",
    "\n",
    "\n",
    "update_after_actions = 4\n",
    "\n",
    "# How often to update the target network\n",
    "target_update_every = 50\n",
    "loss_function = keras.losses.MeanSquaredError()\n",
    "\n",
    "def train(model, env, eps = 0.3, epochs=5):\n",
    "\n",
    "    for epoch in range(epochs):\n",
    "      running_reward = 0\n",
    "      timestep_count = 0\n",
    "\n",
    "      #epsilon = 1\n",
    "      #adpative = True\n",
    "      #if adpative:\n",
    "      #  epsilon = np.exp(-1e-2 * timestep_count)\n",
    "      epsilon = eps#1 - (epoch+1)/epochs\n",
    "      state = env.reset()\n",
    "      while True:\n",
    "          timestep_count += 1\n",
    "\n",
    "          # adpative epsilon decay\n",
    "\n",
    "\n",
    "          # exploration\n",
    "          rand = np.random.uniform(0, 1)\n",
    "          if rand < epsilon:\n",
    "              # Take random action\n",
    "              action = np.random.choice(actsize)\n",
    "          else:\n",
    "              # Predict action Q-values\n",
    "              # From environment state\n",
    "              state_t = tf.convert_to_tensor(state)\n",
    "              state_t = tf.expand_dims(state_t, 0)\n",
    "              action_vals = model(state_t, training=False)\n",
    "              # Choose the best action\n",
    "              action = int(tf.keras.backend.argmax(action_vals))\n",
    "\n",
    "          state_next, reward, done, info = env.step(action)\n",
    "          #state_next = state_next[0]\n",
    "\n",
    "\n",
    "          # Save action/states and other information in replay buffer\n",
    "          action_history.append(action)\n",
    "          state_history.append(state)\n",
    "          state_next_history.append(state_next)\n",
    "          rewards_history.append(reward)\n",
    "          done_history.append(done)\n",
    "          running_rewards_history.append(info['total_reward'])\n",
    "\n",
    "          running_reward = info['total_reward']\n",
    "\n",
    "          state = state_next\n",
    "\n",
    "          # Update every Xth frame to speed up (optional)\n",
    "          # and if you have sufficient history\n",
    "          if timestep_count % update_after_actions == 0 and len(action_history) > batch_size:\n",
    "              # Sample a set of batch_size memories from the history\n",
    "              rand_samples = np.random.randint(0, len(action_history), batch_size)\n",
    "\n",
    "              action_sample = np.array(action_history)[rand_samples]\n",
    "              state_sample = np.array(state_history)[rand_samples]\n",
    "              state_next_sample = np.array(state_next_history)[rand_samples]\n",
    "              rewards_sample = np.array(rewards_history)[rand_samples]\n",
    "              done_sample = np.array(done_history)[rand_samples]\n",
    "\n",
    "              # Create for the sample states the targets (r+gamma * max Q(...)\n",
    "              state_next_sample_t = tf.convert_to_tensor(state_next_sample)\n",
    "              Q_next_state = np.max(model_target.predict(state_next_sample_t), axis=1)   \n",
    "              Q_targets = rewards_sample + gamma * Q_next_state\n",
    "\n",
    "              # What actions are relevant and need updating\n",
    "              relevant_actions = tf.cast(tf.one_hot(action_sample, actsize), tf.float32)\n",
    "\n",
    "              # we will use Gradient tape to do a custom gradient \n",
    "              # in the `with` environment we will record a set of operations\n",
    "              # and then we will take gradients with respect to the trainable parameters\n",
    "              # in the neural network\n",
    "              with tf.GradientTape() as tape:\n",
    "                  # Train the model on your action selecting network\n",
    "                  state_sample_t = tf.convert_to_tensor(np.asarray(state_sample))\n",
    "                  q_values = model(state_sample_t) \n",
    "                  #print(q_values, relevant_actions)\n",
    "                  # We consider only the relevant actions\n",
    "                  Q_of_actions = tf.reduce_sum(tf.multiply(q_values, relevant_actions), axis=1)\n",
    "                  # Calculate loss between principal network and target network\n",
    "                  loss = loss_function(Q_targets, Q_of_actions)\n",
    "\n",
    "                  # Nudge the weights of the trainable variables towards \n",
    "                  grads = tape.gradient(loss, model.trainable_variables)\n",
    "                  optimizer.apply_gradients(zip(grads, model.trainable_variables))\n",
    "\n",
    "          if timestep_count % target_update_every == 0:\n",
    "              # update the the target network with new weights\n",
    "              model_target.set_weights(model.get_weights())\n",
    "              # Log details\n",
    "              template = \"running reward: {:.2f} at episode {}, info: {} epsilon {}\"\n",
    "              print(template.format(running_reward, timestep_count, info, epsilon))\n",
    "\n",
    "          # Don't let the memory grow beyond the limit\n",
    "          if len(rewards_history) > max_memory:\n",
    "              del action_history[:1]\n",
    "              del state_history[:1]\n",
    "              del state_next_history[:1]\n",
    "              del rewards_history[:1]\n",
    "              del done_history[:1]\n",
    "\n",
    "          if done:\n",
    "              print(\"info: \", info)\n",
    "              break\n",
    "      print(\"Epoch \" + str(epoch))\n",
    "      env.render_all()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "321ea504",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "321ea504",
    "outputId": "462c4e26-a19e-4945-dc87-93c58f8528b8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "running reward: 0.91 at episode 50, info: {'total_reward': 0.9091395367703474, 'total_profit': 1.042845055899907, 'position': 1} epsilon 0.3\n",
      "running reward: 1.04 at episode 100, info: {'total_reward': 1.0395263365971437, 'total_profit': 1.2090838808398376, 'position': 0} epsilon 0.3\n",
      "running reward: 0.90 at episode 150, info: {'total_reward': 0.8983930256449256, 'total_profit': 1.2700869042110798, 'position': 1} epsilon 0.3\n",
      "running reward: 1.01 at episode 200, info: {'total_reward': 1.012289400290927, 'total_profit': 1.1967179065564302, 'position': 1} epsilon 0.3\n",
      "running reward: 1.05 at episode 250, info: {'total_reward': 1.0507506067477825, 'total_profit': 1.1598760035535751, 'position': 1} epsilon 0.3\n",
      "running reward: 0.99 at episode 300, info: {'total_reward': 0.9904716399663671, 'total_profit': 1.059251772159015, 'position': 1} epsilon 0.3\n",
      "running reward: 1.31 at episode 350, info: {'total_reward': 1.3084039120513118, 'total_profit': 1.433788853557663, 'position': 1} epsilon 0.3\n",
      "running reward: 1.25 at episode 400, info: {'total_reward': 1.2454247673063237, 'total_profit': 1.3962122584858363, 'position': 1} epsilon 0.3\n",
      "running reward: 1.17 at episode 450, info: {'total_reward': 1.1693851530196882, 'total_profit': 1.2942946654653467, 'position': 0} epsilon 0.3\n",
      "running reward: 1.03 at episode 500, info: {'total_reward': 1.0344574163881062, 'total_profit': 1.1415478657266442, 'position': 1} epsilon 0.3\n",
      "running reward: 1.35 at episode 550, info: {'total_reward': 1.352029223773462, 'total_profit': 1.3246402684910101, 'position': 1} epsilon 0.3\n",
      "running reward: 1.50 at episode 600, info: {'total_reward': 1.5013950671503908, 'total_profit': 1.5252435954327657, 'position': 1} epsilon 0.3\n",
      "running reward: 1.38 at episode 650, info: {'total_reward': 1.3766796543815094, 'total_profit': 1.4345409735852723, 'position': 0} epsilon 0.3\n",
      "running reward: 1.47 at episode 700, info: {'total_reward': 1.4742462480251182, 'total_profit': 1.5288698223607653, 'position': 1} epsilon 0.3\n",
      "running reward: 1.35 at episode 750, info: {'total_reward': 1.3481314312136334, 'total_profit': 1.5181350895234216, 'position': 1} epsilon 0.3\n",
      "running reward: 1.50 at episode 800, info: {'total_reward': 1.4956302600701448, 'total_profit': 1.6097580918813499, 'position': 1} epsilon 0.3\n",
      "running reward: 1.60 at episode 850, info: {'total_reward': 1.5969915762332336, 'total_profit': 1.708371203857506, 'position': 1} epsilon 0.3\n",
      "running reward: 1.67 at episode 900, info: {'total_reward': 1.6731893255300807, 'total_profit': 1.7917138052703692, 'position': 1} epsilon 0.3\n",
      "running reward: 1.78 at episode 950, info: {'total_reward': 1.776202488126687, 'total_profit': 1.924714283138266, 'position': 1} epsilon 0.3\n",
      "info:  {'total_reward': 2.0685405056252635, 'total_profit': 2.1674935347902964, 'position': 0}\n",
      "Epoch 0\n",
      "running reward: 1.12 at episode 50, info: {'total_reward': 1.1182152170092565, 'total_profit': 1.146469830005631, 'position': 1} epsilon 0.3\n",
      "running reward: 1.28 at episode 100, info: {'total_reward': 1.2817847906910886, 'total_profit': 1.3400755853459267, 'position': 0} epsilon 0.3\n",
      "running reward: 1.18 at episode 150, info: {'total_reward': 1.1797159339784675, 'total_profit': 1.2518438026210934, 'position': 1} epsilon 0.3\n",
      "running reward: 1.17 at episode 200, info: {'total_reward': 1.16979976861286, 'total_profit': 1.2837843029213705, 'position': 1} epsilon 0.3\n",
      "running reward: 1.20 at episode 250, info: {'total_reward': 1.2008379450582645, 'total_profit': 1.2371627712767164, 'position': 1} epsilon 0.3\n",
      "running reward: 1.02 at episode 300, info: {'total_reward': 1.0159261610408783, 'total_profit': 1.07914875578261, 'position': 1} epsilon 0.3\n",
      "running reward: 1.37 at episode 350, info: {'total_reward': 1.3652423860639367, 'total_profit': 1.4506851496024065, 'position': 1} epsilon 0.3\n",
      "running reward: 1.36 at episode 400, info: {'total_reward': 1.3597184595219127, 'total_profit': 1.4254322072667136, 'position': 1} epsilon 0.3\n",
      "running reward: 1.34 at episode 450, info: {'total_reward': 1.3442013072136445, 'total_profit': 1.384088967729375, 'position': 1} epsilon 0.3\n",
      "running reward: 1.24 at episode 500, info: {'total_reward': 1.2404291127073734, 'total_profit': 1.2669423208546735, 'position': 1} epsilon 0.3\n",
      "running reward: 1.36 at episode 550, info: {'total_reward': 1.3565001080262857, 'total_profit': 1.3391592806737393, 'position': 0} epsilon 0.3\n",
      "running reward: 1.43 at episode 600, info: {'total_reward': 1.4325425486514807, 'total_profit': 1.4801757473795316, 'position': 1} epsilon 0.3\n",
      "running reward: 1.49 at episode 650, info: {'total_reward': 1.4851888346230402, 'total_profit': 1.4810717812266558, 'position': 0} epsilon 0.3\n",
      "running reward: 1.53 at episode 700, info: {'total_reward': 1.5334095947882316, 'total_profit': 1.5801218598199198, 'position': 1} epsilon 0.3\n",
      "running reward: 1.20 at episode 750, info: {'total_reward': 1.1978095816208152, 'total_profit': 1.415269374660585, 'position': 1} epsilon 0.3\n",
      "running reward: 1.36 at episode 800, info: {'total_reward': 1.3628526785086574, 'total_profit': 1.4492429179503943, 'position': 1} epsilon 0.3\n",
      "running reward: 1.40 at episode 850, info: {'total_reward': 1.404945578665105, 'total_profit': 1.5571145891041809, 'position': 1} epsilon 0.3\n",
      "running reward: 1.30 at episode 900, info: {'total_reward': 1.298081668044219, 'total_profit': 1.5228884425636433, 'position': 0} epsilon 0.3\n",
      "running reward: 1.58 at episode 950, info: {'total_reward': 1.576289357083923, 'total_profit': 1.7875512056328278, 'position': 1} epsilon 0.3\n",
      "info:  {'total_reward': 1.7503383112259914, 'total_profit': 1.973991375421266, 'position': 1}\n",
      "Epoch 1\n",
      "running reward: 0.95 at episode 50, info: {'total_reward': 0.9537473023046584, 'total_profit': 1.0752584624625816, 'position': 1} epsilon 0.3\n",
      "running reward: 1.14 at episode 100, info: {'total_reward': 1.1449091243584513, 'total_profit': 1.2065333977421655, 'position': 1} epsilon 0.3\n",
      "running reward: 1.07 at episode 150, info: {'total_reward': 1.0746831979070586, 'total_profit': 1.200641391238763, 'position': 1} epsilon 0.3\n",
      "running reward: 1.22 at episode 200, info: {'total_reward': 1.2171212032468575, 'total_profit': 1.2665412224794725, 'position': 1} epsilon 0.3\n",
      "running reward: 1.11 at episode 250, info: {'total_reward': 1.1137840295344605, 'total_profit': 1.1611786155588264, 'position': 0} epsilon 0.3\n",
      "running reward: 1.07 at episode 300, info: {'total_reward': 1.0672803679414886, 'total_profit': 1.0984143410024334, 'position': 1} epsilon 0.3\n",
      "running reward: 1.49 at episode 350, info: {'total_reward': 1.4926804498754092, 'total_profit': 1.501548542908391, 'position': 1} epsilon 0.3\n",
      "running reward: 1.43 at episode 400, info: {'total_reward': 1.4260820176796385, 'total_profit': 1.4709704483427861, 'position': 1} epsilon 0.3\n",
      "running reward: 1.36 at episode 450, info: {'total_reward': 1.357316499500131, 'total_profit': 1.4006938859680615, 'position': 1} epsilon 0.3\n",
      "running reward: 1.23 at episode 500, info: {'total_reward': 1.2277122199209962, 'total_profit': 1.2360769804322869, 'position': 0} epsilon 0.3\n",
      "running reward: 1.40 at episode 550, info: {'total_reward': 1.4036245916948964, 'total_profit': 1.3605492256306244, 'position': 1} epsilon 0.3\n",
      "running reward: 1.55 at episode 600, info: {'total_reward': 1.5506796646485272, 'total_profit': 1.4126935218072283, 'position': 1} epsilon 0.3\n",
      "running reward: 1.95 at episode 650, info: {'total_reward': 1.953328205629246, 'total_profit': 1.7015927441139875, 'position': 1} epsilon 0.3\n",
      "running reward: 2.10 at episode 700, info: {'total_reward': 2.098302240132658, 'total_profit': 1.8370703310723278, 'position': 1} epsilon 0.3\n",
      "running reward: 2.01 at episode 750, info: {'total_reward': 2.0084174307347573, 'total_profit': 1.702154675701179, 'position': 1} epsilon 0.3\n",
      "running reward: 2.08 at episode 800, info: {'total_reward': 2.0812358510734814, 'total_profit': 1.8919340990606468, 'position': 0} epsilon 0.3\n",
      "running reward: 2.22 at episode 850, info: {'total_reward': 2.2191541955021923, 'total_profit': 1.9618892049425876, 'position': 1} epsilon 0.3\n",
      "running reward: 2.24 at episode 900, info: {'total_reward': 2.2369597068918337, 'total_profit': 2.1041757457773977, 'position': 1} epsilon 0.3\n",
      "running reward: 2.24 at episode 950, info: {'total_reward': 2.240824563380705, 'total_profit': 2.1540788193446936, 'position': 1} epsilon 0.3\n",
      "info:  {'total_reward': 2.204039770687654, 'total_profit': 2.2266835112339005, 'position': 1}\n",
      "Epoch 2\n",
      "running reward: 1.15 at episode 50, info: {'total_reward': 1.145564907878892, 'total_profit': 1.1384277174722839, 'position': 1} epsilon 0.3\n",
      "running reward: 1.37 at episode 100, info: {'total_reward': 1.3659873730018843, 'total_profit': 1.381220882751959, 'position': 1} epsilon 0.3\n",
      "running reward: 1.19 at episode 150, info: {'total_reward': 1.1891836620937997, 'total_profit': 1.262782224453239, 'position': 1} epsilon 0.3\n",
      "running reward: 1.17 at episode 200, info: {'total_reward': 1.165650467569347, 'total_profit': 1.2773847681597081, 'position': 0} epsilon 0.3\n",
      "running reward: 1.07 at episode 250, info: {'total_reward': 1.0653804443241335, 'total_profit': 1.1584912325019925, 'position': 1} epsilon 0.3\n",
      "running reward: 1.04 at episode 300, info: {'total_reward': 1.03895256942866, 'total_profit': 1.091067192079875, 'position': 1} epsilon 0.3\n",
      "running reward: 1.31 at episode 350, info: {'total_reward': 1.3084888338849976, 'total_profit': 1.4276727335942616, 'position': 1} epsilon 0.3\n",
      "running reward: 1.22 at episode 400, info: {'total_reward': 1.222536614670364, 'total_profit': 1.3678918950185561, 'position': 1} epsilon 0.3\n",
      "running reward: 1.16 at episode 450, info: {'total_reward': 1.1622309674995992, 'total_profit': 1.3169383827586525, 'position': 1} epsilon 0.3\n",
      "running reward: 1.17 at episode 500, info: {'total_reward': 1.1678633985970819, 'total_profit': 1.2374904240792681, 'position': 1} epsilon 0.3\n",
      "running reward: 1.06 at episode 550, info: {'total_reward': 1.063890792242944, 'total_profit': 1.1921933301882024, 'position': 1} epsilon 0.3\n",
      "running reward: 1.30 at episode 600, info: {'total_reward': 1.2987064915462267, 'total_profit': 1.4169505449925386, 'position': 1} epsilon 0.3\n",
      "running reward: 1.16 at episode 650, info: {'total_reward': 1.1599392655362581, 'total_profit': 1.5002698949595044, 'position': 1} epsilon 0.3\n",
      "running reward: 1.23 at episode 700, info: {'total_reward': 1.2324478572844277, 'total_profit': 1.3958255134628739, 'position': 1} epsilon 0.3\n",
      "running reward: 1.00 at episode 750, info: {'total_reward': 1.0024523686428444, 'total_profit': 1.1988548613450927, 'position': 0} epsilon 0.3\n",
      "running reward: 1.15 at episode 800, info: {'total_reward': 1.1541504576107546, 'total_profit': 1.3971359338658706, 'position': 1} epsilon 0.3\n",
      "running reward: 1.21 at episode 850, info: {'total_reward': 1.2132263355152435, 'total_profit': 1.4634217835035435, 'position': 1} epsilon 0.3\n",
      "running reward: 1.19 at episode 900, info: {'total_reward': 1.1918547031238729, 'total_profit': 1.4871225032664606, 'position': 1} epsilon 0.3\n",
      "running reward: 1.31 at episode 950, info: {'total_reward': 1.305691481829097, 'total_profit': 1.6600075317219283, 'position': 1} epsilon 0.3\n",
      "info:  {'total_reward': 1.4013965373653472, 'total_profit': 1.783309489660359, 'position': 1}\n",
      "Epoch 3\n",
      "running reward: 0.90 at episode 50, info: {'total_reward': 0.8965530251453949, 'total_profit': 1.0214625472695362, 'position': 1} epsilon 0.3\n",
      "running reward: 1.08 at episode 100, info: {'total_reward': 1.083186065964254, 'total_profit': 1.2316693809951464, 'position': 1} epsilon 0.3\n",
      "running reward: 1.00 at episode 150, info: {'total_reward': 1.0014461725175186, 'total_profit': 1.256853995947667, 'position': 1} epsilon 0.3\n",
      "running reward: 0.96 at episode 200, info: {'total_reward': 0.9620071016857594, 'total_profit': 1.161975439022377, 'position': 0} epsilon 0.3\n",
      "running reward: 0.88 at episode 250, info: {'total_reward': 0.8840961532902238, 'total_profit': 1.0565699252746714, 'position': 1} epsilon 0.3\n",
      "running reward: 0.78 at episode 300, info: {'total_reward': 0.7790389595205338, 'total_profit': 0.9374759839963317, 'position': 1} epsilon 0.3\n",
      "running reward: 1.01 at episode 350, info: {'total_reward': 1.011039534705082, 'total_profit': 1.256891529176329, 'position': 1} epsilon 0.3\n",
      "running reward: 1.05 at episode 400, info: {'total_reward': 1.0528376094647363, 'total_profit': 1.268413393175708, 'position': 1} epsilon 0.3\n",
      "running reward: 1.08 at episode 450, info: {'total_reward': 1.0755246020879516, 'total_profit': 1.249780907616874, 'position': 1} epsilon 0.3\n",
      "running reward: 1.13 at episode 500, info: {'total_reward': 1.127573876648438, 'total_profit': 1.2100569912622243, 'position': 1} epsilon 0.3\n",
      "running reward: 0.82 at episode 550, info: {'total_reward': 0.8163852998540065, 'total_profit': 1.050880440500521, 'position': 0} epsilon 0.3\n",
      "running reward: 0.90 at episode 600, info: {'total_reward': 0.9025569857838811, 'total_profit': 1.1876354689532518, 'position': 0} epsilon 0.3\n",
      "running reward: 0.95 at episode 650, info: {'total_reward': 0.9513993822990491, 'total_profit': 1.3078498048180207, 'position': 1} epsilon 0.3\n",
      "running reward: 1.06 at episode 700, info: {'total_reward': 1.055675720890082, 'total_profit': 1.230819457955909, 'position': 1} epsilon 0.3\n",
      "running reward: 0.90 at episode 750, info: {'total_reward': 0.897636388776139, 'total_profit': 1.160376984080332, 'position': 1} epsilon 0.3\n",
      "running reward: 1.13 at episode 800, info: {'total_reward': 1.128561819596063, 'total_profit': 1.397451286282374, 'position': 1} epsilon 0.3\n",
      "running reward: 1.04 at episode 850, info: {'total_reward': 1.0385207286414673, 'total_profit': 1.4791475809724395, 'position': 1} epsilon 0.3\n",
      "running reward: 0.97 at episode 900, info: {'total_reward': 0.9749140870545869, 'total_profit': 1.4041705803201507, 'position': 1} epsilon 0.3\n",
      "running reward: 1.03 at episode 950, info: {'total_reward': 1.0322519203055764, 'total_profit': 1.4692204818193926, 'position': 0} epsilon 0.3\n",
      "info:  {'total_reward': 1.13983580803613, 'total_profit': 1.611738332915927, 'position': 1}\n",
      "Epoch 4\n"
     ]
    },
    {
     "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": [
    "# train on Google stock data from timestep 1 to 1000\n",
    "train(model, env_training)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "5575fa1c",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "5575fa1c",
    "outputId": "5e8bf1fe-3cfa-42ff-8622-876de91f4444"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:Compiled the loaded model, but the compiled metrics have yet to be built. `model.compile_metrics` will be empty until you train or evaluate the model.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Found untraced functions such as lstm_cell_layer_call_fn, lstm_cell_layer_call_and_return_conditional_losses, lstm_cell_1_layer_call_fn, lstm_cell_1_layer_call_and_return_conditional_losses while saving (showing 4 of 4). These functions will not be directly callable after loading.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: trained_model/assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: trained_model/assets\n",
      "WARNING:absl:<keras.layers.recurrent.LSTMCell object at 0x7fddc8a5f5d0> has the same name 'LSTMCell' as a built-in Keras object. Consider renaming <class 'keras.layers.recurrent.LSTMCell'> to avoid naming conflicts when loading with `tf.keras.models.load_model`. If renaming is not possible, pass the object in the `custom_objects` parameter of the load function.\n",
      "WARNING:absl:<keras.layers.recurrent.LSTMCell object at 0x7fddc8a5f950> has the same name 'LSTMCell' as a built-in Keras object. Consider renaming <class 'keras.layers.recurrent.LSTMCell'> to avoid naming conflicts when loading with `tf.keras.models.load_model`. If renaming is not possible, pass the object in the `custom_objects` parameter of the load function.\n"
     ]
    }
   ],
   "source": [
    "model.save(\"trained_model\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "f618ca6a",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 312
    },
    "id": "f618ca6a",
    "outputId": "cc557f90-aa4c-4daa-a353-46af73690e88"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Running Reward')"
      ]
     },
     "execution_count": 28,
     "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": [
    "# plot the running rewards\n",
    "plt.plot(running_rewards_history)\n",
    "plt.title(\"Running rewards during training\")\n",
    "plt.xlabel(\"Episodes\")\n",
    "plt.ylabel(\"Running Reward\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "3ff35d9e",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 485
    },
    "id": "3ff35d9e",
    "outputId": "68211704-c8a7-4f5f-8f45-81ead5c29169"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "running reward: 0.94 at episode 50, info: {'total_reward': 0.937765950271371, 'total_profit': 1.0, 'position': 1} epsilon 0\n",
      "running reward: 0.92 at episode 100, info: {'total_reward': 0.9215502253714615, 'total_profit': 1.0, 'position': 1} epsilon 0\n",
      "running reward: 1.13 at episode 150, info: {'total_reward': 1.1334957326386472, 'total_profit': 1.0, 'position': 1} epsilon 0\n",
      "running reward: 1.28 at episode 200, info: {'total_reward': 1.2842121995337292, 'total_profit': 1.0, 'position': 1} epsilon 0\n",
      "running reward: 1.14 at episode 250, info: {'total_reward': 1.1364569287360526, 'total_profit': 1.0, 'position': 1} epsilon 0\n",
      "running reward: 1.28 at episode 300, info: {'total_reward': 1.2773663798752863, 'total_profit': 1.0, 'position': 1} epsilon 0\n",
      "running reward: 1.23 at episode 350, info: {'total_reward': 1.2260555968814382, 'total_profit': 1.0, 'position': 1} epsilon 0\n",
      "running reward: 1.10 at episode 400, info: {'total_reward': 1.1007441486821272, 'total_profit': 1.0, 'position': 1} epsilon 0\n",
      "running reward: 1.18 at episode 450, info: {'total_reward': 1.1804858179041584, 'total_profit': 1.0, 'position': 1} epsilon 0\n",
      "info:  {'total_reward': 1.1440764754166266, 'total_profit': 1.1824518836696618, 'position': 1}\n",
      "Epoch 0\n"
     ]
    },
    {
     "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": [
    "# test model results on the next leg of the stock data from time 1000 to 1500\n",
    "train(model, env_1, 0, 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "79b1a1ad",
   "metadata": {
    "id": "79b1a1ad"
   },
   "source": [
    "*The reward of the Deep Q Learning function after training for 1000 timesteps is: 186.441452 and the total profit is 1.5480215108859194. The initial money increased by 155%.*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "72ab0674",
   "metadata": {
    "id": "72ab0674"
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "include_colab_link": true,
   "name": "RL - Stock Trading - Custom Reward.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}