ReinforcementLearning.jl

ReinforcementLearning.jl copied to clipboard

A call to state(env::TicTacToeEnv, Observation{BitArray{3}}()) does not result in the correct style.

The error can be "seen" using:

env = TicTacToeEnv()
display(state(env, Observation{String}(), current_player(env))) # works as expected
display(state(env, Observation{String}()))                      # works as expected
display(state(env, Observation{BitArray{3}}(), current_player(env))) # works as expected
display(state(env, Observation{BitArray{3}}())) # reports state default style: Observation{Int}

A simple fix is to the following line to TicTacToeEnv.jl

RLBase.state(env::TicTacToeEnv, ::Observation{BitArray{3}}) = env.board