Open nimzodisaster opened 3 years ago
I think this is a problem similar to mine. I would like to do a multivariate distance matrix regression using a random variable. However, the random variable is not hierarchical. Using the lme4 package would be possible, but not using matrix. Would it be possible to implement? Thanks
I am interested in expanding MDMR functionality to also use lme from nlme because lme (as oppsed to lmer) allows for specification of variance-covariance correlation structures in the mixed model.
Are you aware of any gotchas re the MDMR method, before attempting this modification?