a-paxton/paramtests

Automated exploration of mixed-effects model parameters in R

★ 0Forks 0RGitHub ↗Compare

README

paramtests: Automated exploration of mixed-effects model parameters in R

This package facilitates explorations of mixed-effects models by automatically running all permutations of the various parameters. The functions currently only address linear mixed-effects models, although mixed logit models are in development. Note the dependencies required for each of the functions in the package, listed below in "Dependencies."

Current functions

  • mrest (Maximal Random Effects: Slope Testing): For linear mixed-effects models. This function runs all permutations of random effects structures and outputs resulting model fit statistics OR warnings produced. This function uses fixed dependent variables, predictors, and random intercepts.
  • omse (Optimal Model Structure Exploration): For linear mixed-effects models. This function runs all permutations of fixed and random effects model structures and outputs resulting model fit statistics OR warnings produced. This function uses fixed dependent variables and random intercepts.

Dependencies

paramtests functions require the following packages to run:

  • mrest: gtools, lme4, plyr
  • omse: gtools, lme4, plyr

Installation

This package may be downloaded and installed or may be directly installed from R.

Downloading and installing manually from R

  1. Download ZIP of paramtests.
  2. Unzip file and change folder name to paramtests (from paramtests-master).
  3. Move folder to target location.
  4. In R, change working directory to new parent folder for "paramtests":
    setwd("/parent-folder-here/")
    
  5. Install.
    install("paramtests")
    
  6. Confirm installation.
    ?paramtests
    ?omse
    ?mrest
    

Installing from GitHub in R

  1. Install (if necessary) and load devtools package.
    install.packages("devtools")
    library(devtools)
    
  2. Use install_github() function to install directly.
    install_github("a-paxton/paramtests",dependencies=TRUE)
    

Contributors

a-paxton

Issues