search

PacktPublishing / Deep-Reinforcement-Learning-Hands-On

Hands-on Deep Reinforcement Learning, published by Packt
MIT License
2.85k stars 1.29k forks source link

bug on indexing? #8

Closed TomLin closed 5 years ago

TomLin commented 6 years ago

https://github.com/PacktPublishing/Deep-Reinforcement-Learning-Hands-On/blob/a307a952bb914a1b5a43b7c92b7237ca28d88d28/Chapter14/06_train_d4pg.py#L79-L86

Hello Maxim,

Your book is really helpful for us to immediately implement D4PG, but I am wondering if the index for l and u should be modified to be as follows:

proj_distr[eq_dones, l[eq_mask]] = 1.0

proj_distr[ne_dones, l[ne_mask]] = (u - b_j)[ne_mask] proj_distr[ne_dones, u[ne_mask]] = (b_j - l)[ne_mask]

Thanks for any clarification.

Shmuma commented 6 years ago

Hi!

Yes, you're right, those indexations are now needed for code to work properly, but I'm not sure why it worked fine before :). Probably some subtle change in numpy indexation. I'll implement and commit the change you've proposed.

Thanks!

Shmuma commented 5 years ago

Sorry for looong wait. Fixed by PR13: https://github.com/PacktPublishing/Deep-Reinforcement-Learning-Hands-On/pull/13