mergeR2: Merge partR2 objects to combine R2s for main effects and...

View source: R/mergeR2.R

mergeR2R Documentation

Merge partR2 objects to combine R2s for main effects and interactions

Description

The function merges partR2 object based on a full model with interactions with a partR2 object based on a reduced model without interaction. The reduced model is used to infer main effect semi-partial R2s. This function essentially takes over the complete partR2 object for the full model and adds semi-partial R2s which have been calculated based on the reduced model and are not already present in the full model partR2 object (which can be main effects). The function also combines the bootstrap estimates, accessible with partR2_obj$R2_boot.

Usage

mergeR2(R2_full, ...)

Arguments

R2_full

partR2 object for the full model, with the interaction (but not the main effects) in partvars.

...

other partR2 objects, which do not contain the interaction so that the semi-partial R2s for the main effects could be calculated.

Details

This function is experimental and should be used with caution. See vignette or paper on how to use it to obtain semi-partial R2s for main effects which are also involved in interactions.

Value

Returns an object of class partR2, which takes most components from the full model except for semi-partial R2s.

Examples


data(biomass)
library(lme4)

# scale data
biomass[] <- lapply(biomass, function(x) if (is.double(x)) scale(x) else x)

# Full model
mod_full <- lmer(Biomass ~  Year + Temperature * Precipitation + SpeciesDiversity + (1|Population),
            data = biomass)
# Semi-partial R2 for interaction and all other predictors of interest
(R2_full <- partR2(mod_full, partvars = c("Temperature:Precipitation", "SpeciesDiversity", "Year"),
data = biomass))

# model without interaction to get main effect semi-partial R2s
mod_noIA <- lmer(Biomass ~  Year + Temperature + Precipitation + SpeciesDiversity + (1|Population),
            data = biomass)
(R2_noIA <- partR2(mod_noIA, partvars = c("Temperature", "Precipitation"), data = biomass))

# combine both
(R2_comb <- mergeR2(R2_full, R2_noIA))


partR2 documentation built on May 29, 2024, 2:29 a.m.