lcc
Nov 25, 2023
·
1 min read
lcc estimates longitudinal concordance correlation (lcc), longitudinal Pearson correlation (lpc), and longitudinal accuracy (la) by fitting polynomial mixed-effects regression models. It implements the variance-components approach of Oliveira, Hinde & Zocchi (2018) and provides inference, confidence intervals (parametric and bootstrap), and graphical summaries for agreement profiles over time.
Key features
- Handles balanced or unbalanced designs and allows covariates in the linear predictor.
- Models heteroscedastic within-group errors with or without time covariates.
- Generates fitted values, sampled values, and confidence bands numerically and graphically.
Package details
- Version: 2.0.0 (CRAN; archived 2025-11-07 pending fixes; development continues on GitHub).
- Depends on R ≥ 3.2.3 plus nlme, ggplot2; imports hnp, parallel, doSNOW, doRNG, foreach.
- Suggests roxygen2, covr, testthat, MASS.
- Maintainer: Thiago de Paula Oliveira thiago.paula.oliveira@alumni.usp.br.
- Authors: Thiago de Paula Oliveira (aut/cre), Rafael de Andrade Moral, Silvio Sandoval Zocchi, Clarice Garcia Borges Demetrio, John Hinde.
- License: GPL (≥ 2).
Installation:
install.packages("lcc") # CRAN archive
# Development
install.packages("devtools")
devtools::install_github("Prof-ThiagoOliveira/lcc")
A companion Shiny app illustrates how parameters affect LCC over time, and tutorials are available via the PeerJ paper (10.7717/peerj.9850). Former CRAN binaries/sources remain in the archive while the next stable release is prepared.
