I maintain several packages for R, including some lightly modified versions of existing packages. You can see all of them on my [github]{https://github.com/davidaarmstrong} page. Below are some links to the ones I think are more prominent or useful to a general audience.
ClarkeTest
The clarkeTest
package implements Kevin Clarke’s (2007) distribution-free test for non-nested models. It does so in a way that is more modular than the games
package currently supporting models estimated with lm()
, glm()
(binomial
, poisson
and negbin
links), polr()
(from the MASS
package), clm()
(from the ordinal
package), multinom()
(from the nnet
package) and mlogit()
(from the mlogit
package)
DAMisc
The DAMisc
package has evolved over the past decade to include many of the functions I use when teaching applied stats to social scientists. I think many of the functions might be useful more broadly so I thought it would be worth discussing the functionality in a sort of thematic way here. The functions do fall into a few different themes.
- Functions that attempt to figure out whether and what kind of unmodeled non-linearities exist.
- Functions for investigating interactions in different settings (linear models and binomial GLMs).
- Functions for post-model evaluation and examination of non-linear models (GLMs, ordinal data models and unordered data models).
factorplot
The factorplot
package provides users with a way to visualize pairwise comparisons that arise from multiple situations. The package was first developed to plot pairwise comparisons that arise from factors being included in statistical models; that is, to deal with the reference category problem. It has grown to include multinomial logit models (estimated with mlogit()
from nnet
), eff
objects created with the effects
package and glht
objects from the multcomp
package. The link to the pkgdown
site above provides some examples for various different use-cases. For more information about methods for visualizing paired comparisons, see my article in the R Journal.