adriantorrie / dqn-ddpg_stock_trading Goto Github PK

View Code? Open in Web Editor NEW

Using DQN/DDPG for stock trading. Xiong, Z., Liu, X.Y., Zhong, S., Yang, H. and Walid, A., 2018. Practical deep reinforcement learning approach for stock trading, NeurIPS 2018 AI in Finance Workshop.

License: MIT License

Jupyter Notebook 24.73% Python 50.23% Dockerfile 0.03% HTML 25.01%

dqn-ddpg_stock_trading's Introduction

Practical Deep Reinforcement Learning Approach for Stock Trading

Prerequisites

Python 3.6 envrionment

CMake, OpenMPI

Installation of system packages CMake, OpenMPI on Mac

brew install cmake openmpi

Activate your envrionment using using conda or Anaconda

source activate myenv

Install gym under this environment

pip install gym

Download Official Baseline Package

Clone the repo and cd into it:

git clone https://github.com/openai/baselines.git
cd baselines

Install tendorflow
```
pip install Tensorflow
```
should be sufficient. Refer to TensorFlow installation guide for more details.
Install baselines package
```
pip install -e .
```

Testing the installation

All unit tests in baselines can be run using pytest runner:

pip install pytest
pytest

Replace files with files in this repository and change file address

gym

Find your gym package under environment folder, in my computer it is under

/Users/xiongzhuoran/anaconda3/envs/venv/lib/python3.6/site-packages/gym/

Replece the file

gym\envs\__init__.py

with file from this repository

DQN_Stock_Trading/gym/envs/__init__.py

Add folder in this repository to gym\envs in your computer

DQN_Stock_Trading/gym/envs/zxstock of this repository

Open

gym/envs/zxstock/zxstock_env.py and gym/envs/zxstock/zxstock_testenv.py

change the address at line 9 and line 10 into where you want to save the image

Baseline

Open your baselines folder cloned before, find

baselines/baselines/run.py

Replace it with

DQN_Stock_Trading/baselines/baselines/run.py in this reposotory

Training model and Testing

If you only want to train the model run this

python -m baselines.run --alg=ddpg --env=ZXStock-v0 --network=mlp --num_timesteps=1e4

If you also want to see the testing result

python -m baselines.run --alg=ddpg --env=ZXStock-v0 --network=mlp --num_timesteps=1e4 --play

Some Other Commands May Need:

Tensorflow Update

pip install --upgrade tensorflow==1.11.0

pip3 install opencv-python
pip3 install lockfile
pip3 install -U numpy
pip3 install mujoco-py==0.5.7

Please cite the following paper

Xiong, Z., Liu, X.Y., Zhong, S., Yang, H. and Walid, A., 2018. Practical deep reinforcement learning approach for stock trading, NeurIPS 2018 AI in Finance Workshop.

Recommend Projects