[R-pkgs] Bayesian Fractional Polynomials package "bfp" on CRAN

Daniel Sabanés Bové daniel.sabanesbove at ifspm.uzh.ch
Mon Sep 27 11:44:50 CEST 2010


  Fractional polynomials ("FPs") are an automatic way of fitting 
non-linear, parametric effects. The R-package mfp implements a 
frequentist inference approach for FP models. Recently, we have proposed 
a Bayesian inference approach for normal FP models, which is based on 
the quasi-default hyper-/g/ prior for the regression coefficients [1]. 
This approach is implemented in the new R-package "bfp".

The R-package bfp (current version: 0.0-17) is now available on CRAN [2].
The current development version is (still) available on R-Forge [3].

For a quick start try:

install.packages("bfp")
library(bfp)
example(BmaSamples)

and if you have more time, to reproduce results from the paper [1]:

demo(ozone)

Questions, suggestions or critique are very welcome!

Best regards,
Daniel

[1] Statistics & Computing paper: 
http://dx.doi.org/10.1007/s11222-010-9170-7
[2] CRAN: http://cran.r-project.org/web/packages/bfp/index.html
[3] R-Forge: http://r-forge.r-project.org/projects/bfp/



More information about the R-packages mailing list