Plotting GAMM interactions with ggplot2 is not always easy, especially if you have some interactions in there . Plotting the results of GAMMs are not always so easy . I’ll be using the diamonds dataset for this, as it has some nice non-linear features .

I’m working a lot with Generalized Additive Mixed Mixed Models (GAMMs) I started seeing some differences in the output from other mixed models I had run through [lme4] and even Bayesian mixed models with [rstanarm] The random effects terms were specified differently . At first I thought this was mainly the fact that I was now using smoothing splines, which I was unfamiliar with running before .

A co-worker was working on some generalized additive mixed models (GAMM) through the R package [mgcv] The analyses work very well, and results were as expected . There were also some built in plotting functions for the gamm output in R . However, they were with base-r, and we all know base r is not the most beautiful .