DRL4Trading DRL4Trading is an open source Python framework for building, training and evaluating trading algorithms using deep reinforcement learning. This framework uses numpy, pandas, tensorflow (v2) and tf-agents. Features and Limitations - Add your own indicator. Users can add any calculated feature to the data by implementing a simple calss. For example a technical indicator or a fundamental parameter can be added by implementing "CustomColumn" class as seen in "preprocessdata/customcolumns.py". - Multiple csv files. You can use several csv files for one stock. for example you can load "price data" for each ticker (open, high, low, close ...) in a directory and "fundamental data" in another directory. Just add the list of dirs in settings file (config/settins.py) in section "CSVFILESETTINGS->dirlist". - Only csv files. As of today, the framework work just with csv data. you should provide separate csv files for every single ticker. - Only PPO. as of today, this framework only works with PPO algorithm. more models will be added later if I can find more free time. - Supports DTL.

⚠ No verified equity curve — no track-record source connected.

No score history is stored yet — only the current score is shown.

⚠ No live verification account connected — ask for proof before buying.

Alerts on changes: coming soon

Prop-firm compatibility not provided.

Data-completeness & trust index (not a profitability rating)