I have been enjoying fitting RL model using pymc4 so far! However I just now encountered an issue that I have no idea how to fix. In my study, I aim to compare two RL models. RL1 is just the classic Q learning model. RL2 is pretty much the same except I added another parameter that dictates how my t…

I see I see. I need them as a placeholder of a trial because my model builds in a forgetting mechanism. If there isn’t any convenient aesara function, this workaround I think worked: idx_noNaN_t, idx_noNaN_b, idx_noNaN_s = at.nonzero(~at.isnan(rewards_)) utility_ = rewards_.copy() utility_ = at.set…

Divided by zero encountered when fitting a RL model

Questions v5

ricardoV94 July 20, 2022, 6:06am 2

I doubt your likelihood will handle nan out of the box, if you didn’t do anything specific tp account for them.

Can you show the complete code that reproduces the problem you are observing?

Topic		Replies	Views
Reinfocement learning model - derivative of RV is zero Questions	5	1787	April 10, 2019
Problems when fitting a reinforcement learning model to data Questions modeling	4	541	April 20, 2022
Divergence encountered when fitting RL model v5 modeling	2	542	July 30, 2022
Sampling crashes with "The derivative of RV [whatever] is zero" when using a custom likelihood defined with pm.Potential Questions	4	4028	February 23, 2019
Fitting a complex model with a large dataset: chain contains only diverging samples Questions	5	1276	December 12, 2019

Divided by zero encountered when fitting a RL model

Related topics