urgently need your help! In doing bayesian modelling, how do we set prior for the standard deviation of regression slope, so-called hyper-parameter. For instance, y = b0 + b1x1+b2x2+error, where b2 ~ N(0, tau), and tau ~ U(1,1). so, how to set the prior for the tau? Many thanks!

In the following, I modified the example codes in ?rstan so that the modified codes use a hyper parameter as follows Y \sim \text{Normal}(\mu,1), \mu \sim \text{Normal}(hyper, 10), hyper \sim \text{Normal}(-1,1) So, we can describe the prior of hyper parameter in the ususal way. stanmodelcode <…

Many thanks @Jean_Billie ! Is it possible to set those in the function ‘brm’? I would like to set prior for hyper-parameter the brm. Actually I am not familiar with rstan syntax. Thanks.

Sorry, I cannot. I am not familiar with brm package, but only Stan.

You can do this in brms by passing Stan code to brm() via two arguments: prior and stanvars (which adds tau as a parameter). Effectively this is exactly what @Jean_Billie recommends, but leans on brms to create as much of the Stan code as possible. bpriors <- prior(normal(0, 2), class = "Intercept"…

@Jean_Billie Thanks a lot for your kind reply.

How to set prior for hyper-parameter

Interfaces brms

ecorenhb May 23, 2020, 9:10am 7

@wpetry Many thanks! It is very helpful! I will try those.

Topic		Replies	Views
Hierarchical prior for the regression coefficients brms	3	1332	November 2, 2021
Specifying a hierarchical prior on mixed model with brms brms	3	2341	May 24, 2019
Brms mulitlevel mode: syntax issue Modeling rstan , specification , brms	2	101	July 24, 2024
How to set prior on var in brms brms techniques , brms	2	175	June 7, 2025
How to interpret estimated posterior distribution of hyperparameters returned by brms brms brms	7	1037	June 8, 2021

How to set prior for hyper-parameter

Related topics