How does one obtain the covariance matrix for the fixed effects from fitting a mixed effect model using fitlme or fitlmematrix?

Is there a way to easily obtain the covariance matrix for the fixed effects estimates from a linear mixed effects model fit by fitlme or fitlmematrix?

I'm fitting the linear mixed effects model

lme= fitlme(tbl,['SpO2~SaO2*ITA+ (SaO2|Subj)'])

I would like to get the covariance matrix for the estimates of fixed effects SaO2*ITA. Is there a way to det?

Matlab gives you the covariance matrix for the random effects via psi = covarianceParameters(lme), but I don't want to think about how to use it to compute the covariance matrix for the fixed effects. The function p = coefTest(lme,H,C) must compute the fixed effects covariance matrix when it tests the contrast H in the fixed effects at hypothesized value C, but does not have a name-value argument that would provide the fixed effects covariance matrix.

Thanks for any help you can give.

Answer 1

[the cyclist]

A LinearMixedModel object lme has a CoefficientCovariance property, which you can obtain with

lme.CoefficientCovariance

This is cited in the documentation I linked. It is the second property in the list.

[Gene Pennello]

Thank you! Somehow I didn't see this property. I appreciate it very much.