0xflotus / piecewisesem

Piecewise Structural Equation Modeling in R

piecewiseSEM: Piecewise Structural Equation Modeling in R

Version 2.1.0

Last updated: 05 December 2019

To install

Run the following code to install the development version:

install.packages("piecewiseSEM")

Getting Help

See our website at (piecewiseSEM)[http://jslefche.github.io/piecewiseSEM/]

There is an online resource available for SEM, including piecewiseSEM and lavaan, available (here)[https://jslefche.github.io/sem_book/]

This version is a major update to the piecewiseSEM package that uses a completely revised syntax that better reproduces the base R syntax and output. It is highly recommended that consult vignette("piecewiseSEM") even if you have used the package before as it documents the many changes.

It also incorporates new functionality in the form of coefficient standardization and updated methods for R^2 for mixed models.

Currently supported model classes: lm, glm, gls, pgls, sarlm, lme, glmmPQL, lmerMod, merModLmerTest, glmerMod

To add a new model class

Frequently, we get requests to add new model classes. We'd like to accommodate wherever we can! Currently, to add a new model class, you will need to update the following functions in the following files:

• In coefs.R - getCoefficients() - will need a method to generate a standardized coefficient table from your class - GetSDy() - will need a method to get the proper SD of y, particularly for non-Gaussian error families
• In helpers.R - GetSingleData() - need a method to get a data frame from your class type - GetOLRE() - if your class has random effects, will need a method to get observation level random effects - nObs() - if your class has a non-standard way to get the number of observations, will need it here. - You might need to write a all.vars method for you class. See all.vars.merMod() as an example.
• In psem.R - evaluateClasses() - so we know that this class can work in piecewiseSEM
• In rsquared.R - rsquared() - what function should be used to get the rsquared for your model class? - Will need an rsquared.youclass() method to get the R^2

Example

# Install development branch from github
library(devtools)
install_github("jslefche/piecewiseSEM@devel", build_vignette = TRUE)

# Load library
library(piecewiseSEM)

# Read vignette
vignette("piecewiseSEM")

# Create fake data
set.seed(1)

data <- data.frame(
  x = runif(100),
  y1 = runif(100),
  y2 = rpois(100, 1),
  y3 = runif(100)
)

# Store in SEM list
modelList <- psem(
  lm(y1 ~ x, data),
  glm(y2 ~ x, "poisson", data),
  lm(y3 ~ y1 + y2, data),
  data
)

# Run summary
summary(modelList)

# Address conflict using conserve = T
summary(modelList, conserve = T)

# Address conflict using direction = c()
summary(modelList, direction = c("y2 <- y1"))

# Address conflict using correlated errors
modelList2 <- update(modelList, y2 %~~% y1)

summary(modelList2)