Projection pursuit with monotone scoring for ordered categorical (ordinal) data in R, together with the replication materials for the article ordpp: Projection Pursuit with Monotone Scoring for Ordinal Data in R (D. A. Perez Ruiz, submitted to The R Journal).
From CRAN (once available):
install.packages("ordpp")
Development version from this repository (the package lives in the ordpp/ folder):
# install.packages("remotes")
remotes::install_github("GITHUB-USERNAME/ordpp", subdir = "ordpp")
library(ordpp)
sim <- simulate_ordpp(n = 600, p = 6, seed = 103)
fit <- ordpp(sim$x[1:300, ], scoring = "item", seed = 105)
fit
heldout_test_ordpp(fit, sim$x[301:600, ], B = 999, seed = 106)
ordpp/ the R package (GPL-3)
scripts/ replication scripts, run from the repository root
results/, figures/ outputs used by the article
perez-ruiz.tex article; RJwrapper.tex is the R Journal wrapper
ordpp.bib, packages.bib
R CMD INSTALL ordpp
install.packages(c("psychTools", "psych", "Gifi")) # in R; ordPens from the CRAN archive
Rscript scripts/reproduce_article.R # every number, table and figure; a few minutes
FULL=1 Rscript scripts/reproduce_article.R # also re-runs the long simulations (hours)
pdflatex RJwrapper; bibtex RJwrapper; pdflatex RJwrapper; pdflatex RJwrapper; pdflatex RJwrapper
The SAPA (bfi) data are loaded from the psychTools package and are not
redistributed here.