SurrogateRsq: Goodness-of-Fit Analysis for Categorical Data using the Surrogate R-Squared

To assess and compare the models' goodness of fit, R-squared is one of the most popular measures. For categorical data analysis, however, no universally adopted R-squared measure can resemble the ordinary least square (OLS) R-squared for linear models with continuous data. This package implement the surrogate R-squared measure for categorical data analysis, which is proposed in the study of Dungang Liu, Xiaorui Zhu, Brandon Greenwell, and Zewei Lin (2022) <doi:10.1111/bmsp.12289>. It can generate a point or interval measure of the surrogate R-squared. It can also provide a ranking measure of the percentage contribution of each variable to the overall surrogate R-squared. This ranking assessment allows one to check the importance of each variable in terms of their explained variance. This package can be jointly used with other existing R packages for variable selection and model diagnostics in the model-building process.

Version: 0.2.1
Depends: R (≥ 3.5.0), MASS (≥ 7.3-54), PAsso (≥ 0.1.10), progress (≥ 1.2.0), scales (≥ 1.1.1)
Suggests: R.rsp, knitr, rmarkdown, testthat (≥ 3.0.0), dplyr (≥ 1.1.1)
Published: 2023-04-24
Author: Xiaorui (Jeremy) Zhu [aut, cre, cph], Dungang Liu [ctb], Zewei Lin [ctb], Brandon Greenwell [ctb]
Maintainer: Xiaorui (Jeremy) Zhu <zhuxiaorui1989 at>
License: GPL-2 | GPL-3 [expanded from: GPL (≥ 2)]
NeedsCompilation: no
Materials: README NEWS
CRAN checks: SurrogateRsq results


Reference manual: SurrogateRsq.pdf
Vignettes: A introduction to categorical data goodness-of-fit analysis using the SurrogateRsq package


Package source: SurrogateRsq_0.2.1.tar.gz
Windows binaries: r-prerel:, r-release:, r-oldrel:
macOS binaries: r-prerel (arm64): SurrogateRsq_0.2.1.tgz, r-release (arm64): SurrogateRsq_0.2.1.tgz, r-oldrel (arm64): SurrogateRsq_0.2.1.tgz, r-prerel (x86_64): SurrogateRsq_0.2.1.tgz, r-release (x86_64): SurrogateRsq_0.2.1.tgz
Old sources: SurrogateRsq archive


Please use the canonical form to link to this page.