Any way to fix some random variables during inference?

[lithces]

Hi,

I am wondering if it is possible to fix some variables during the inference process.
e.g. in Regression,

• we already know some sigma for some outputs,
• we already know some coefficients (weights) of the regression

I am asking this because I am trying to learn a model with pylds and some elements in the transition matrix is already known.

Thanks !