Hello all, Is it possible to directly expose the gradient of the log-likelihood in Stan for every transition and to define it as a variable in the model? I saw that the last mention of this was more than 4 years ago: Expose gradient ope…

Thanks for the answer! So how would one access the gradient or calculate (and store) it on the fly? Is that even possible?

You would have to either do it analytically and / or write your own C++ (very carefully).

+1. If you do it analytically or numerically, you can do it in Stan. If you’re willing to get to C++, anything is possible.

ok thanks! Is there an example how to analytically or numerically calculate the gradient in Stan? C++ is over my head atm (unless it is easier to implement there). Thanks again!

and based on this discussion: [image] How to print gradient in .stan for debug RStan I know print(“lp=”, target()) can print out log-posterior for debug. How can I print gradient? Thank you. Ziyi Chen is it still possible? Because I didn’t see the gradient info for…

It is in the columns that start with grad, although that is the gradient with respect to the unconstrained parameters, which may not be what you want.

I checked my *diagnostic.csv again and I have columns that start with g_* for different parameters. Is this the gradient in the unconstrained space? (I don’t see columns that start with grad.) Also, what does p_* for different parameters mean?

[image] nerpa: Is this the gradient in the unconstrained space? yes [image] nerpa: Also, what does p_* for different parameters mean? momentum

Ok, thanks! That’s super helpful. Now is there a way to access the gradient values for each iteration as I sample my model? I don’t need the values for each leapfrog step, so the unconstrained space and per-iteration gradient value is exactly what I need.

Exposing the gradient of the log-likelihood

General

nerpa October 28, 2020, 10:41am 7

and based on this discussion:

is it still possible? Because I didn’t see the gradient info for each iteration as one of the outputs in the diagnostic file from CmdStan.

Topic		Replies	Views
Use gradient of log likelihood in model block General	2	442	February 14, 2021
Using Stan to compute the gradient of a function General	15	806	June 1, 2024
How to print gradient in .stan for debug RStan	4	871	November 2, 2018
Log_prob_grad via csv files for cmdstan CmdStan	7	487	May 12, 2021
A gradient block? General	19	2386	August 11, 2021

Exposing the gradient of the log-likelihood

Related topics