Stepwise covariate modeling (SCM) for nonlinear mixed-effects
models fitted with 'nlmixr2'. Forward inclusion and backward elimination
are driven by likelihood-ratio tests, and the covariate terms are
generated inside the model body, so continuous covariates are centered
and categorical covariates expanded into indicator columns without
editing the model by hand. Candidate fits can be cached and resumed,
fitted in parallel, and reviewed through per-step and all-candidate
summary tables. The approach follows Jonsson and Karlsson (1998)
<doi:10.1023/A:1011970125687>, and the implementation in
'Perl-speaks-NONMEM' described by Lindbom, Ribbing and Jonsson (2004)
<doi:10.1016/j.cmpb.2003.11.003>.
| Version: |
0.4 |
| Depends: |
R (≥ 4.1) |
| Imports: |
checkmate, cli (≥ 3.4.0), data.table, lotri, nlme, nlmixr2est (≥ 7.0.0), nlmixr2utils (≥ 0.3), rxode2 (≥ 5.0.0), stats, tools, utils |
| Suggests: |
future, knitr, nlmixr2data, pkgload, rmarkdown, testthat (≥
3.0.0), withr |
| Published: |
2026-09-24 |
| DOI: |
10.32614/CRAN.package.nlmixr2scm (may not be active yet) |
| Author: |
Justin Wilkins 
[aut, cre, cph],
Matthew Fidler
[aut],
Yaping Liu [aut],
Bill Denney [aut],
Vipul Mann [aut],
Vishal Sarsani
[aut],
Christian Bartels [ctb] |
| Maintainer: |
Justin Wilkins <justin.wilkins at occams.com> |
| BugReports: |
https://github.com/nlmixr2/nlmixr2scm/issues |
| License: |
GPL (≥ 3) |
| URL: |
https://github.com/nlmixr2/nlmixr2scm |
| NeedsCompilation: |
no |
| Materials: |
README, NEWS |
| CRAN checks: |
nlmixr2scm results |