ReinforcementLearning.jl

ReinforcementLearning.jl copied to clipboard

Here goes a fix and associated unit test for issue "Wrong style for state report for TicTacToeEnv() #1079"

hespanha

Wrong style for state report for TicTacToeEnv()

2

A call to `state(env::TicTacToeEnv, Observation{BitArray{3}}())` does not result in the correct style. The error can be "seen" using: ```julia env = TicTacToeEnv() display(state(env, Observation{String}(), current_player(env))) # works as expected display(state(env,...

hespanha

Needless allocations in reward() and is_terminated() for

2

The two functions reward(::TicTacToeEnv,::Player) s_terminated(::TicTacToeEnv) result in a small but needless allocation due to a type instability in call to `get_tic_tac_toe_state_info()` To see this, you can use: ```julia using ReinforcementLearning...

hespanha

Does it allow defining an environment that has continuous action space? And how?

4

First of all, I would like to say thank you to all of the contributors of this useful package! I am a learner of both RL and this package. I...

WuSiren

Fix for TicTacToeEnv allows illegal moves #1001

2

TicTacToeEnv allows one state to be played multiple times by the same agent. This condition should prevent this behavior and throw an error in case the state is played multiple...

SanteriVtj

PR Checklist - [ ] Update NEWS.md? - [ ] Unit tests for all structs / functions? - [ ] Integration and correctness tests using a simple env? - [...

jeremiahpslewis

Design of run loop and hooks

8

I am running into limitations of the current design of the `run` loop: Let's assume I am using a custom policy that internally stores the history of past observations and...

johannes-fischer