Fixed overflow issue of the while loop inside `predict_next_evaluation`

LucasAlegre / morl-baselines

Multi-Objective Reinforcement Learning algorithms implementations.

https://lucasalegre.github.io/morl-baselines

MIT License

295 stars 47 forks source link

Fixed overflow issue of the while loop inside `predict_next_evaluation` #33

Closed ChrisZonghaoLi closed 1 year ago

ChrisZonghaoLi commented 1 year ago

Now added a flag to stop the iteration if current_neighb_threshold goes infinity and causes overflow trouble.

ChrisZonghaoLi commented 1 year ago

Yes, I did! When it fails to find at least 4 neighbors it will raise the error and stop the simulations

LucasAlegre commented 1 year ago

Hi, it doesn't seem right to me to raise ValueError in case the algorithm should end. It should stop without crashing everything. In which case will this ValueError be raised exactly? Do you need a specific combination of hyperparamters or it can happen anytime? @ChrisZonghaoLi @ffelten