Untitled2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e08ccd57-8bc9-42f2-954e-e675c61d74a9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-01-16 00:00:00 -> 2017-01-20 23:55:00 length: 1440\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import gym\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "from stable_baselines3.common.env_checker import check_env\n",
    "from stable_baselines3 import PPO, TD3\n",
    "from stable_baselines3.common.vec_env import DummyVecEnv, SubprocVecEnv,VecNormalize\n",
    "from stable_baselines3.common.env_util import make_vec_env\n",
    "from stable_baselines3.common.utils import set_random_seed\n",
    "from neo_finrl.env_fx_trading.env_fx import tgym\n",
    "\n",
    "env_name =\"TradingGym-v0\"\n",
    "env = gym.make(env_name)\n",
    "check_env(env)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "f050376e-e3f1-40e6-b700-a094b43c1cb5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "./data/split/GBPUSD/weekly/GBPUSD_2017_0.csv\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-01-02 01:00:00 -> 2017-01-06 23:55:00 length: 1428\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-01-02 01:00:00 -> 2017-01-06 23:55:00 length: 1428\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-01-02 01:00:00 -> 2017-01-06 23:55:00 length: 1428\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-01-02 01:00:00 -> 2017-01-06 23:55:00 length: 1428\n",
      "new model: ppo_fx-2017-4\n",
      "./data/split/GBPUSD/weekly/GBPUSD_2017_1.csv\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-01-09 00:00:00 -> 2017-01-13 23:55:00 length: 1440\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-01-09 00:00:00 -> 2017-01-13 23:55:00 length: 1440\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-01-09 00:00:00 -> 2017-01-13 23:55:00 length: 1440\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-01-09 00:00:00 -> 2017-01-13 23:55:00 length: 1440\n"
     ]
    },
    {
     "ename": "NameError",
     "evalue": "name 'kwargs' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m~\\AppData\\Local\\Temp/ipykernel_14288/2143296813.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m     32\u001b[0m         \u001b[0mdf\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpd\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mread_csv\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfile\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     33\u001b[0m         \u001b[0mtotal_timesteps\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m200000\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 34\u001b[1;33m         \u001b[0mtrain\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdf\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mtotal_timesteps\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mmodel_name\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     35\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Temp/ipykernel_14288/2143296813.py\u001b[0m in \u001b[0;36mtrain\u001b[1;34m(df, total_timesteps, model_name, num_cpu, **policy_kwargs)\u001b[0m\n\u001b[0;32m     13\u001b[0m     \u001b[0menv\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mVecNormalize\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0menv\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnorm_obs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mTrue\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnorm_reward\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mTrue\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mclip_obs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m10.\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     14\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0mos\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpath\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mexists\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mmodel_name\u001b[0m \u001b[1;33m+\u001b[0m \u001b[1;34m'.zip'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 15\u001b[1;33m         \u001b[0mmodel\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mPPO\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mload\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mmodel_name\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0menv\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0menv\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     16\u001b[0m         \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34mf'retaining model: {model_name}'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     17\u001b[0m         \u001b[0mmodel\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlearn\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtotal_timesteps\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mtotal_timesteps\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mNameError\u001b[0m: name 'kwargs' is not defined"
     ]
    }
   ],
   "source": [
    "\n",
    "env_name =\"TradingGym-v0\"\n",
    "def train(df, total_timesteps,model_name, num_cpu = 4, **policy_kwargs):\n",
    "    # num_cpu Number of processes to use\n",
    "    # Create the vectorized environment\n",
    "    # env = SubprocVecEnv([make_env(env_id, i) for i in range(num_cpu)])\n",
    "\n",
    "    # Stable Baselines provides you with make_vec_env() helper\n",
    "    # which does exactly the previous steps for you.\n",
    "    # You can choose between `DummyVecEnv` (usually faster) and `SubprocVecEnv`\n",
    "    # env = make_vec_env(t, n_envs=num_cpu, seed=0, vec_env_cls=DummyVecEnv)\n",
    "    env = DummyVecEnv([lambda: tgym(df=df) for _ in range(num_cpu)] )\n",
    "    # Automatically normalize the input features and reward\n",
    "    env = VecNormalize(env, norm_obs=True, norm_reward=True, clip_obs=10.)\n",
    "    if os.path.exists(model_name + '.zip'):\n",
    "        model = PPO.load(model_name, env=env, **kwargs)\n",
    "        print(f'retaining model: {model_name}')\n",
    "        model.learn(total_timesteps=total_timesteps)\n",
    "        model.save(model_name)\n",
    "        env.save(\"vec_normalize.pkl\")\n",
    "    else:\n",
    "        print(f'new model: {model_name}')\n",
    "        model = PPO('MlpPolicy', env=env, **policy_kwargs)\n",
    "        model.learn(total_timesteps=total_timesteps)\n",
    "        model.save(model_name)   \n",
    "\n",
    "model_name = \"ppo_fx-2017-4\"   \n",
    "for i in range(0,52): # one year 2017\n",
    "    \n",
    "    file = f\"./data/split/GBPUSD/weekly/GBPUSD_2017_{i}.csv\"\n",
    "    if os.path.isfile(file):\n",
    "        print(file)\n",
    "        df = pd.read_csv(file)\n",
    "        total_timesteps=200000\n",
    "        train(df,total_timesteps,model_name)\n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f454a883-9c62-42bf-84cc-0df0e59a9e9a",
   "metadata": {},
   "outputs": [],
   "source": [
    "import gym\n",
    "import numpy as np\n",
    "\n",
    "from stable_baselines3 import PPO\n",
    "from stable_baselines3.common.vec_env import DummyVecEnv, SubprocVecEnv\n",
    "from stable_baselines3.common.env_util import make_vec_env\n",
    "from stable_baselines3.common.utils import set_random_seed\n",
    "model = PPO.load(\"test\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8e36ffa5-9a34-4c26-bb53-f4e3e0239de9",
   "metadata": {},
   "outputs": [],
   "source": [
    "import gym\n",
    "from stable_baselines3.common.env_checker import check_env\n",
    "from neo_finrl.env_fx_trading.env_fx import tgym\n",
    "env_name =\"TradingGym-v1\"\n",
    "env = gym.make(env_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "56f1b516-121e-41f6-8105-7026768a907f",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a434fb76-08cc-420e-8758-e6163956c202",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "./data/split/GBPUSD/weekly/GBPUSD_2017_0.csv\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-01-02 01:00:00 -> 2017-01-06 23:55:00 length: 1428\n",
      "retaining model: td3_fx-2017\n",
      "./data/split/GBPUSD/weekly/GBPUSD_2017_1.csv\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-01-09 00:00:00 -> 2017-01-13 23:55:00 length: 1440\n",
      "retaining model: td3_fx-2017\n",
      "./data/split/GBPUSD/weekly/GBPUSD_2017_2.csv\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-01-16 00:00:00 -> 2017-01-20 23:55:00 length: 1440\n",
      "retaining model: td3_fx-2017\n",
      "./data/split/GBPUSD/weekly/GBPUSD_2017_3.csv\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-01-23 00:00:00 -> 2017-01-27 23:55:00 length: 1440\n",
      "retaining model: td3_fx-2017\n",
      "./data/split/GBPUSD/weekly/GBPUSD_2017_4.csv\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-01-30 00:00:00 -> 2017-02-03 23:55:00 length: 1440\n",
      "retaining model: td3_fx-2017\n",
      "./data/split/GBPUSD/weekly/GBPUSD_2017_5.csv\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-02-06 00:00:00 -> 2017-02-10 23:55:00 length: 1440\n",
      "retaining model: td3_fx-2017\n",
      "./data/split/GBPUSD/weekly/GBPUSD_2017_6.csv\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-02-13 00:00:00 -> 2017-02-17 23:55:00 length: 1440\n",
      "retaining model: td3_fx-2017\n",
      "./data/split/GBPUSD/weekly/GBPUSD_2017_7.csv\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-02-20 00:00:00 -> 2017-02-24 23:55:00 length: 1440\n",
      "retaining model: td3_fx-2017\n",
      "./data/split/GBPUSD/weekly/GBPUSD_2017_8.csv\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-02-27 00:00:00 -> 2017-03-03 23:55:00 length: 1440\n",
      "retaining model: td3_fx-2017\n",
      "./data/split/GBPUSD/weekly/GBPUSD_2017_9.csv\n",
      "initial done:\n",
      "observation_list:['Open', 'High', 'Low', 'Close', 'minute', 'hour', 'day', 'macd', 'boll_ub', 'boll_lb', 'rsi_30', 'dx_30', 'close_30_sma', 'close_60_sma']\n",
      " assets:['GBPUSD']\n",
      " time serial: 2017-03-06 00:00:00 -> 2017-03-10 23:55:00 length: 1440\n",
      "retaining model: td3_fx-2017\n"
     ]
    }
   ],
   "source": [
    "\n",
    "env_name =\"TradingGym-v0\"\n",
    "def train(df, total_timesteps,model_name, num_cpu = 4, **policy_kwargs):\n",
    "    # num_cpu Number of processes to use\n",
    "    # Create the vectorized environment\n",
    "    # env = SubprocVecEnv([make_env(env_id, i) for i in range(num_cpu)])\n",
    "\n",
    "    # Stable Baselines provides you with make_vec_env() helper\n",
    "    # which does exactly the previous steps for you.\n",
    "    # You can choose between `DummyVecEnv` (usually faster) and `SubprocVecEnv`\n",
    "    # env = make_vec_env(t, n_envs=num_cpu, seed=0, vec_env_cls=DummyVecEnv)\n",
    "    env = DummyVecEnv([lambda: tgym(df=df)] )\n",
    "    \n",
    "    # Automatically normalize the input features and reward\n",
    "    env = VecNormalize(env, norm_obs=True, norm_reward=True, clip_obs=10.)\n",
    "    if os.path.exists(model_name + '.zip'):\n",
    "        model = TD3.load(model_name, env=env, **policy_kwargs)\n",
    "        print(f'retaining model: {model_name}')\n",
    "        model.learn(total_timesteps=total_timesteps)\n",
    "        model.save(model_name)\n",
    "        env.save(\"vec_normalize.pkl\")\n",
    "    else:\n",
    "        print(f'new model: {model_name}')\n",
    "        model = TD3('MlpPolicy', env=env, **policy_kwargs)\n",
    "        model.learn(total_timesteps=total_timesteps)\n",
    "        model.save(model_name)   \n",
    "        \n",
    "\n",
    "model_name = \"td3_fx-2017\"   \n",
    "for i in range(0,52): # one year 2017\n",
    "    \n",
    "    file = f\"./data/split/GBPUSD/weekly/GBPUSD_2017_{i}.csv\"\n",
    "    if os.path.isfile(file):\n",
    "        print(file)\n",
    "        df = pd.read_csv(file)\n",
    "        total_timesteps=2000\n",
    "        policy_kwargs ={}\n",
    "        train(df,total_timesteps,model_name, **policy_kwargs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "51346e7e-3466-4a07-a4db-69e7e165e919",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.6rc1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}