ReinforcementLearning.jl

ReinforcementLearning.jl copied to clipboard

The definition for MultiThreadEnv exists under src/algorithms/policy_gradient/multi_thread_env.jl but that file is not included in policy_gradient.jl , algorithms.jl, or ReinforcementLearningZoo.jl. Is this intentional?

After trying to include it in policy_gradient.jl, I ran into some dependency issues with MacroTools and EnrichedAction not being defined. I think I solved MacroTools by adding using MacroTools and adding it to ReinforcementLearningZoo/Project.toml.

EnrichedAction appears to be defined in the archived ReinforcementLearningZoo.jl repository (here). Should this also be defined in the ReinforcementLearningZoo package in this repo?

Is this intentional?

yes, policy gradient related algorithms are not ready yet in the master branch.

You may read the source code at https://github.com/JuliaReinforcementLearning/ReinforcementLearning.jl/tree/v0.10.1 for the stable version.