Nonlinear mixed models

Licensed under the Creative Commons attribution-noncommercial license. Please share & remix noncommercially, mentioning its origin.

Nonlinear mixed models

• Gaussian response (i.e., not generalized)

nlme package: advantages

• well-established, stable
• reasonably well-documented (Pinheiro and Bates 2000)
• flexible heteroscedasticity models (weights argument)
• residual correlation/R-side effects (correlation argument)

nlme: disadvantages

• Gaussian responses only
• relatively slow
• crossed effects are hard

lme4::nlmer

• in principle way better
• nonlin function needs derivatives provided
• fixed effects are hard to incorporate (need to write the function yourself, or hack the output of derivs)

Pinheiro, José C., and Douglas M. Bates. 2000. Mixed-Effects Models in S and S-PLUS. New York: Springer.