search

yuriak / RLQuant

Applying Reinforcement Learning in Quantitative Trading
343 stars 102 forks source link

The train_reward is always nan. #2

Closed bingzh8801 closed 6 years ago

bingzh8801 commented 6 years ago

Hello, When I run PairsTradingTutorial.ipynb to trade, the train_reward is always nan, as follows:

train_reward: nan
train_reward: nan
train_reward: nan
train_reward: nan
...

I couldn't find what's wrong with the code,could you please help me? By the way, I wonder the version of packages we used is different, What versions of packages are you using now ?

Thank you very much.

yuriak commented 6 years ago

@bingzh8801 Sorry for the delay, I've just noticed this issue and have replied you via email.