ugr-sail / sinergym

Gym environment for building simulation and control using reinforcement learning
https://ugr-sail.github.io/sinergym/
MIT License
127 stars 34 forks source link

(v3.3.3) - Sinergym extra variables logging #412

Closed AlejandroCN7 closed 4 months ago

AlejandroCN7 commented 4 months ago

Description

This PR enables Sinergym to log variables that are not part of the observation space. This is ideal for obtaining building information while not interfering with or complicating the learning process of DRL agents.

Motivation and Context

The operating method involves having environments with all variables and then applying a wrapper that removes the desired ones from the observation space, but in such a way that their log is included. A notebook has also been created that explains the use case to use it correctly.

Fixes #410

Types of changes

Checklist:

Changelog:

Extra: