Machine Learning with TensorFlow/Chapter 8. Reinforcement learning.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Chapter 8. Reinforcement learning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from googlefinance.client import get_price_data, get_prices_data, get_prices_time_data\n",
    "            \n",
    "from matplotlib import pyplot as plt         \n",
    "import numpy as np                           \n",
    "import tensorflow as tf                      \n",
    "import random"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_prices(prices):\n",
    "    \n",
    "    fix, ax = plt.subplots(figsize=(10,10))\n",
    "    \n",
    "    plt.title('Opening stock prices')\n",
    "    plt.xlabel('day')\n",
    "    plt.ylabel('price ($)')\n",
    "    ax.plot(prices)\n",
    "    \n",
    "    plt.tight_layout()\n",
    "    plt.savefig('prices.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_prices(param, cache_filename='stock_pricesa.npy'):\n",
    "    try:                                                                   \n",
    "        stock_prices = np.load(cache_filename)\n",
    "    except IOError:\n",
    "        stock_hist = get_price_data(param)\n",
    "        stock_prices = stock_hist.Open \n",
    "        np.save(cache_filename, stock_prices.values)                              \n",
    "\n",
    "    return stock_prices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {},
   "outputs": [],
   "source": [
    "param = {\n",
    "\t'q': \".DJI\", # Stock symbol (ex: \"AAPL\")\n",
    "\t'i': \"86400\", # Interval size in seconds (\"86400\" = 1 day intervals)\n",
    "\t'x': \"INDEXDJX\", # Stock exchange symbol on which stock is traded (ex: \"NASD\")\n",
    "\t'p': \"15Y\" # Period (Ex: \"1Y\" = 1 year)\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "prices = get_prices(param)\n",
    "plot_prices(prices)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {},
   "outputs": [],
   "source": [
    "class DecisionPolicy:\n",
    "    def select_action(self, current_state):                 \n",
    "        pass\n",
    "\n",
    "    def update_q(self, state, action, reward, next_state):  \n",
    "        pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {},
   "outputs": [],
   "source": [
    "class RandomDecisionPolicy(DecisionPolicy):           \n",
    "    def __init__(self, actions):\n",
    "        self.actions = actions\n",
    "\n",
    "    def select_action(self, current_state):          \n",
    "        action = random.choice(self.actions)\n",
    "        return action"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [],
   "source": [
    "def run_simulation(policy, initial_budget, initial_num_stocks, prices, hist):\n",
    "    budget = initial_budget                                                \n",
    "    num_stocks = initial_num_stocks                                        \n",
    "    share_value = 0                                                        \n",
    "    transitions = list()\n",
    "    \n",
    "    for i in range(len(prices) - hist - 1):\n",
    "        if i % 1000 == 0:\n",
    "            print('progress {:.2f}%'.format(float(100*i) / (len(prices) - hist - 1)))\n",
    "        current_state = np.asmatrix(np.hstack((prices[i:i+hist], budget, num_stocks)))                                                         \n",
    "        current_portfolio = budget + num_stocks * share_value              \n",
    "        action = policy.select_action(current_state)                    \n",
    "        share_value = float(prices[i + hist])\n",
    "        \n",
    "        if action == 'Buy' and budget >= share_value:                      \n",
    "            budget -= share_value\n",
    "            num_stocks += 1\n",
    "        elif action == 'Sell' and num_stocks > 0:                          \n",
    "            budget += share_value\n",
    "            num_stocks -= 1\n",
    "        else:                                                              \n",
    "            action = 'Hold'\n",
    "        new_portfolio = budget + num_stocks * share_value                  \n",
    "        reward = new_portfolio - current_portfolio                         \n",
    "        next_state = np.asmatrix(np.hstack((prices[i+1:i+hist+1], budget, num_stocks)))\n",
    "        transitions.append((current_state, action, reward, next_state))\n",
    "        policy.update_q(current_state, action, reward, next_state)         \n",
    "\n",
    "    portfolio = budget + num_stocks * share_value                          \n",
    "    return portfolio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [],
   "source": [
    "def run_simulations(policy, budget, num_stocks, prices, hist):\n",
    "    num_tries = 10                                                         \n",
    "    final_portfolios = list()                                              \n",
    "    for i in range(num_tries):\n",
    "        final_portfolio = run_simulation(policy, budget, num_stocks, prices, hist)                                                                 \n",
    "        final_portfolios.append(final_portfolio)\n",
    "        print('Final portfolio: ${}'.format(final_portfolio))\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "progress 0.00%\n",
      "progress 26.51%\n",
      "progress 53.02%\n",
      "progress 79.53%\n",
      "Final portfolio: $144636.77999999962\n",
      "progress 0.00%\n",
      "progress 26.51%\n",
      "progress 53.02%\n",
      "progress 79.53%\n",
      "Final portfolio: $153730.45\n",
      "progress 0.00%\n",
      "progress 26.51%\n",
      "progress 53.02%\n",
      "progress 79.53%\n",
      "Final portfolio: $199721.61999999988\n",
      "progress 0.00%\n",
      "progress 26.51%\n",
      "progress 53.02%\n",
      "progress 79.53%\n",
      "Final portfolio: $133901.62999999983\n",
      "progress 0.00%\n",
      "progress 26.51%\n",
      "progress 53.02%\n",
      "progress 79.53%\n",
      "Final portfolio: $151355.69999999995\n",
      "progress 0.00%\n",
      "progress 26.51%\n",
      "progress 53.02%\n",
      "progress 79.53%\n",
      "Final portfolio: $155358.45999999996\n",
      "progress 0.00%\n",
      "progress 26.51%\n",
      "progress 53.02%\n",
      "progress 79.53%\n",
      "Final portfolio: $163360.39999999976\n",
      "progress 0.00%\n",
      "progress 26.51%\n",
      "progress 53.02%\n",
      "progress 79.53%\n",
      "Final portfolio: $147395.77000000005\n",
      "progress 0.00%\n",
      "progress 26.51%\n",
      "progress 53.02%\n",
      "progress 79.53%\n",
      "Final portfolio: $186565.74000000025\n",
      "progress 0.00%\n",
      "progress 26.51%\n",
      "progress 53.02%\n",
      "progress 79.53%\n",
      "Final portfolio: $187326.54000000004\n"
     ]
    }
   ],
   "source": [
    "prices = get_prices(param)\n",
    "# plot_prices(prices)\n",
    "\n",
    "actions = ['Buy', 'Sell', 'Hold']                               \n",
    "hist = 3\n",
    "\n",
    "policy = RandomDecisionPolicy(actions)                          \n",
    "budget = 100000.0                                               \n",
    "num_stocks = 0                                                  \n",
    "run_simulations(policy, budget, num_stocks, prices, hist)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}