partR2

'partR2' is an R package to quantify semi-partial R2 in linear mixed-effect models, providing a tool for decomposing the coefficient of determination (R2) into the variance explained by specific predictors. This package offers a unique approach by iteratively removing predictors of interest and assessing the change in the variance of the linear predictor, thereby quantifying the amount of variance uniquely explained by each predictor or set of predictors.

Additionally, 'partR2' estimates structure coefficients, representing the correlation between a predictor and fitted values, providing an independent measure of a fixed effect's overall contribution to prediction. The package also includes the calculation of inclusive R2, beta weights, and implements parametric bootstrapping for estimating confidence intervals.