Advance Topics: New section for "RL for Trading" #558
Conversation
|
As suggested, have created this PR to include a section for "RL for Trading" in advance topics. Please review and let me know for any changes. Thanks. |
| ### Create an environment with custom parameters | ||
|
|
||
| You can change parameters such as dataset, frame_bound, etc while creating the environment. | ||
| To try out and explore, you can use two default datasets available in the GitHub repository - [*FOREX*](https://github.com/AminHP/gym-anytrading/blob/master/gym_anytrading/datasets/data/FOREX_EURUSD_1H_ASK.csv) and [*Stocks*](https://github.com/AminHP/gym-anytrading/blob/master/gym_anytrading/datasets/data/STOCKS_GOOGL.csv), but you can use your own as well. |
There was a problem hiding this comment.
If using your own dataset requires having some pre-defined columns, it may be helpful to elaborate here.
There was a problem hiding this comment.
Checked their documentation and there is no mention of pre-defined columns for custom environments. Assume, the two sample datasets that are provided should be used as references.
Slightly wary of including this to avoid mentioning anything beyond documentation.
Please share your thoughts on this.
There was a problem hiding this comment.
Yes, I think you're right. I looked into the source-code too, and it seems that there are no special columns needed for df. It's up to the user to extend TradingEnv and use the columns they want.
|
|
||
| # This is only for illustration. | ||
| # Values such as state_space, stock_dimension should be derived in the notebook before using them here. | ||
| env_kwargs = { |
There was a problem hiding this comment.
Isn't is simpler to give some mock (or default) values for all variables?
There was a problem hiding this comment.
Have made this clearer, hope this works.
| Create a new environment using df_test to simulate test scenario. | ||
|
|
||
| ```python | ||
| trained_ppo = PPO.load("trained_models/agent_ppo.zip") |
There was a problem hiding this comment.
PPO is not imported anywhere in the notebook. I guess it can be retrieved e.g. through finrl.stablebaselines3.models.DRLAgent.get_model("ppo").
There was a problem hiding this comment.
It is imported through stablebaseline3, have included a import statement.
|
|
||
| df_account_value_ppo, df_actions_ppo = DRLAgent.DRL_prediction( | ||
| model=trained_ppo, | ||
| environment = e_test) |
There was a problem hiding this comment.
Thanks, have updated this now.
| Their [tutorial series](https://finrl.readthedocs.io/en/latest/tutorial/Guide.html) is very exhaustive and beginner friendly. One can choose to start at various levels depending on familiarity (introduction, advance, practical, optimization, others). | ||
|
|
||
|
|
||
| ## Additional readings |
There was a problem hiding this comment.
There is also this Gymnasium environment gym-trading-env that might be worth mentioning in this section (cf. https://gym-trading-env.readthedocs.io/, linked also on Gymnasium docs under the Third-party Environments section https://gymnasium.farama.org/environments/third_party_environments/#gym-trading-env-trading-environment).
There was a problem hiding this comment.
Thanks for your suggestions, have included a link in additional readings.
|
Hi @dantp-ai , thanks for the detail reviews. I am slightly held up with another PR, will check your comments, make changes and re-submit by next weekend. Thanks again !! |
|
Thanks @dantp-ai for your reviews !! Just wanted to check if this is good to be merged and go live. |
|
Hope you are having a great day !! Was wondering if there is anything pending from my side. If you have been pondering on the topic relevance and whether to include this into the course material, understand. Please let me know either way. |
|
LGTM. Nothing from my side. |
closes #555