pwSEM: Piecewise Structural Equation Modelling

Conduct dsep tests (piecewise SEM) of a directed, or mixed, acyclic graph without latent variables (but possibly with implicitly marginalized or conditioned latent variables that create dependent errors) based on linear, generalized linear, or additive modelswith or without a nesting structure for the data. Also included are functions to do desp tests step-by-step,exploratory path analysis, and Monte Carlo X2 probabilities. This package accompanies Shipley, B, (2026).Cause and Correlation in Biology: A User's Guide to Path Analysis, StructuralEquations and Causal Inference (3rd edition). Cambridge University Press.

Version:	1.0.0
Depends:	R (≥ 2.10)
Imports:	copula, gamm4, ggm, igraph, mgcv, poolr, stats
Suggests:	knitr, rmarkdown, testthat (≥ 3.0.0)
Published:	2025-07-02
DOI:	10.32614/CRAN.package.pwSEM
Author:	Bill Shipley [aut, cre]
Maintainer:	Bill Shipley <Bill.Shipley at USherbrooke.ca>
License:	MIT + file LICENSE
NeedsCompilation:	no
Materials:	README, NEWS
CRAN checks:	pwSEM results

Documentation:

Reference manual:	pwSEM.html , pwSEM.pdf
Vignettes:	pwSEM (source, R code)

Downloads:

Package source:	pwSEM_1.0.0.tar.gz
Windows binaries:	r-devel: pwSEM_1.0.0.zip, r-release: pwSEM_1.0.0.zip, r-oldrel: pwSEM_1.0.0.zip
macOS binaries:	r-release (arm64): pwSEM_1.0.0.tgz, r-oldrel (arm64): pwSEM_1.0.0.tgz, r-release (x86_64): pwSEM_1.0.0.tgz, r-oldrel (x86_64): pwSEM_1.0.0.tgz

Linking:

Please use the canonical form https://CRAN.R-project.org/package=pwSEM to link to this page.