mfp2: Multivariable Fractional Polynomial Models with Extensions

Fits multivariable fractional polynomial models with selection of predictors and functional forms. The mfp2() function supports generalized linear models, including Gamma and inverse-Gaussian regression; negative-binomial, multinomial and ordinal regression; generalized estimating equations for clustered data; Cox and parametric survival models; and Fine-Gray competing-risks models. Extensions include approximate cumulative distribution transformations for sigmoid-shaped associations and spike-at-zero modelling for semi-continuous covariates. The mfpi() function assesses interactions between a categorical grouping variable and continuous, binary, or categorical predictors. Fractional polynomial modelling is described by Royston and Altman (1994) <doi:10.2307/2986270> and Royston and Sauerbrei (2008, ISBN:978-0-470-02842-1). The approximate cumulative distribution extension follows Royston (2014) <doi:10.1177/1536867X1401400206> and Royston and Sauerbrei (2016) <doi:10.1177/1536867X1601600109>. Spike-at-zero modelling follows Becher et al. (2012) <doi:10.1002/bimj.201100263>. Interaction analyses follow Royston and Sauerbrei (2004) <doi:10.1002/sim.1815>, Royston and Sauerbrei (2013) <doi:10.1002/sim.5813>, and Royston and Sauerbrei (2014) <doi:10.1002/sim.6308>.

Package details

AuthorEdwin Kipruto [aut, cre], Michael Kammer [aut], Patrick Royston [aut], Willi Sauerbrei [aut]
MaintainerEdwin Kipruto <kiprutoedwin2626@gmail.com>
LicenseGPL-3
Version1.1.1
URL https://github.com/EdwinKipruto/mfp2 
Package repositoryView on CRAN
Installation Install the latest version of this package by entering the following in R:
install.packages("mfp2")

Try the mfp2 package in your browser

Any scripts or data that you put into this service are public.

mfp2 documentation built on Nov. 14, 2026, 1:06 a.m.