chz056/bear

bear's Issues

Update Notice

Update Notice - BEAR v1.79

Dear BEAR users,

We are excited to announce several significant updates to BEAR, the Physics-Principled Building Environment for Control and Reinforcement Learning.

New Features:

• Improved User Interface: We have introduced the get_user_info() function to gather user inputs. This function offers an interactive and user-friendly experience to customize simulation parameters.
• Customizable Reward Function: The reward function is now customizable, offering you the flexibility to design the reward to fit your specific research goals. This can be easily done through get_user_info() function as well.
• Rewrite of the quickstart.py script: To improve readability and usage, we've rewritten the quickstart.py script to demonstrate how to initialize and use the environment.
• Enhanced error detection and input validation: Our updated system is now more robust against incorrect or invalid user inputs.

We hope these updates make BEAR more useful and user-friendly for your research. We look forward to your feedback and encourage you to report any issues you may encounter.

Best regards,
The BEAR Team

Ways of customizing reward function

Hi, thanks for developing this framework. I am curious whether there is a tutorial for customizing reward function: from the paper, it seems that the trade-off parameter beta can be easily tuned but is there a way to design reward functions with totally different format? Thanks!!!