search

fabsig / GPBoost

Combining tree-boosting with Gaussian process and mixed effects models
Other
571 stars 46 forks source link

How to get the model variance for a random intercept model using gpboost? #128

Closed raheems closed 9 months ago

raheems commented 9 months ago

I fitted a random intercept model with one grouping variable. I get the model summary such as

Model summary:
 Log-lik    AIC    BIC
 -181.17 388.34 451.42
Nb. observations: 946
Nb. groups: 113 (id)
-----------------------------------------------------
Covariance parameters (random effects):
     Param.
id  24.6481

How to get the model variance in addition to the covariance of the random effects? In statsmodels, there is a Scale parameter which shows the model variance.

fabsig commented 9 months ago

What's your likelihood? It seems you have a non-Gaussian likelihood such as a binary. If you have a binary likelihood, there is no other parameter. If you have likelihood = "gaussian", you will see the error variance.