Parallel fitting does not apply to DDM models with reinforcement learning

[St-ZFeng]

Greetings! After many confirmations, I found that the chains parameter is invalid for models such as HDDMrl. This will get:

PicklingError: Could not pickle the task to send it to the workers.

Is this feature not yet implemented in the RL model? Or am I using the wrong fitting parameters?

Thank you for your help and this handy docker version that saves me from the bad experience of the original version.

[panwanke]

Greetings! After many confirmations, I found that the chains parameter is invalid for models such as HDDMrl. This will get:

PicklingError: Could not pickle the task to send it to the workers.

Is this feature not yet implemented in the RL model? Or am I using the wrong fitting parameters?

Thank you for your help and this handy docker version that saves me from the bad experience of the original version.

Thank you for your feedback. Currently RL cannot support parallel multiple MCMC chains because RLDDM was not developed by the original authors of HDDM. And, it also failed after we tried to fix this issue.

However, we do allow RLHDDM to run multiple chains with a single core, just add, parallel=False to the parameter.

Here is a example notebook RLDDM.ipynb