• Updated contact email
  • Fixed issue with S3 methods and all
  • Now supports model classes: glmmTMB , Sarlm
  • Added d-sep and Chi-squared tests for gam models
  • Added support for tibble and data.table
  • Added stop_psem to deal with unsupported functions, e.g., poly
  • Fixed issue with infinite loops with sortDag
  • Fixed issue with getSDx not being able to get numeric from data.frame with additional attributes
  • Fixed issue with GetData obtaining transformed data from lme4 objects
  • Fixed issue with non-recursive models and igraph::is_dag
  • Fixed issue with GetSDy not returning correlation when NAs are present
  • Fixed issue with arguments in formulae for all.vars_trans and all.vars_notrans
  • Fixed issue with GetSDy and glm.nb
  • Fixed issue with rsquared when random slopes were not present as corresponding fixed effects
  • Fixed issue with GetSingleData getting transformed data from gam models
  • Fixed issue with anovaLRT returning weird rownames
  • Fixed issue passing single object to anova.psem
  • Added log-likelihood chi-squared goodness-of-fit and AIC statistics from Shipley & Douma 2020
  • Added coefficient standardization for GLM(M)
  • Added preliminary support for GAMs
  • Fixed issue with interactions coded with “:” and coefs
  • Fixed issue with AICc and AIC.type = "dsep" calculating incorrectly
  • Updated coefs to remove depracated function emmeans::CLD
  • Fixed bug with interactions without main effects in coefs
  • Fixed issue with rsquared.glm for “gaussian” family andmethod = "nagelkerke" -Fixed issue withcar::Anovaoutput andhandleCategoricalCoefs`
  • implemented igraph package for DAG construction and evaluation
  • replaced pbkrtest with car::Anova
  • added plot.psem function using DiagrammR package
  • coefs now returns coefficients for categorical variables using emmeans and Anova
  • added multigroup function and analysis
  • added sqrt link for poisson GLMMs
  • Changed contact e-mail to new affiliation
  • Fixed bug with KRp and lmerMod models with intercepts
  • Fixed bug with getOLRE and single observation-level random effects glmerMods
  • Fixed errant parentheses in rsquared leading to wrong values for lme
  • Fixed bug in cyclic where incorrect error was returned
  • Fixed bug with partialCorr and negative correlations returning wrong P-value
  • Fixed with coefs and standardization with mixed models
  • Fixed bug with KRp and uneven sample size
  • New warning issued when NAs present in the dataset
  • Added Gamma distribution to rsquared
  • Fixed bug to determine whether graph is cyclic
  • Fixed bug thowing errors with calls to lmer
  • Fixed bug creating errors with objects fit using lmerTest
  • Added pkgdown website

New syntax

  • All functions have been re-written from the ground up
  • Incorporates new psem function and S3 objects
  • All necessary information can now be obtained with a single function summary

Updated R[2] functions

  • Extends to new distributions and model types using a single function rsquared

New standardization procedures

  • Implements range standardization for all response types
  • Adds multiple forms of standardization for binary responses, see ?coefs

NOTES

  • Removed sem.plot function
  • Removed sem.lavaan function
  • groups= argument is currently broken but will be fixed in version 2.1 (see doc)