Backtest when fast_train=false raises error

ZhengyaoJiang / PGPortfolio

PGPortfolio: Policy Gradient Portfolio, the source code of "A Deep Reinforcement Learning Framework for the Financial Portfolio Management Problem"(https://arxiv.org/pdf/1706.10059.pdf).

GNU General Public License v3.0

1.74k stars 750 forks source link

Backtest when fast_train=false raises error #36

Open astanziola opened 6 years ago

astanziola commented 6 years ago

Greetings! First of all, thank you for the amazing work! I find it a very elegant method and the paper is great at describing it.

I think I've encountered an issue while doing backtest with the flag fast_train=false in the settings. It looks like the function _evaluate(), at line 80 of tradertrainer.py, is called with the argument set_name='validation', but is not able to handle such input. Is it something which is still to be developed?

Thank you!

ZhengyaoJiang commented 6 years ago

The fast_train flag seems to be out of date. I will check that later.