Plot(hypothesis) with multivariate model

Interfaces brms
1

Hi, I want to get a graph that shows the prior and posterior of my parameters and I think that’s what I can get with plot(hypothesis).

plot(hypothesis(model, class = “b”, ‘TaskDif_cwc > 0’))

However, when I use the formula, I always get the error message:

Error: Some parameters cannot be found in the model: ‘b_TaskDif_cwc’

I tried this with many different parameters. Could this be because I have a multivariate model? How can I specify the response variable? I have tried to add resp= but that didn’t help. Also note that I have specified a prior so it’s not the flat default of brms for regression coefficients.

2

Hi Pia,

if you gave a multivariate model then the parameters are often prefixed with each model’s y, i.e., if you have a multivariate model such as:

mvbind(A,B,C) ~ examples

where examples contains categorical variables G and H to estimate, then using summary() you should, I believe, see A_exampleG, B_exampleG, ..., C_exampleH

3

Brilliant! That was it. Thanks for your great help as always!

1 Like