devtools::install_github("jdonhauser/MicEnvMod")
Variable importance for a glm by permuting predictor variables
Plot variable importance from randomForest as barplot with ggplot
Plot response a dependent variable to a particular predictor. Calculated by setting all predictor variables except one to a fixed value.
Bivariate response plot, same as respMono, but with two predictor variables.
model cross validation by repeated split sampling. Currently models of class randomForest and glm are supported.
glm cross validation by repeated split sampling with stepwise selection. Assesses stability of predictor variable selection.
Cross validation of weighte ensemble models from different model types by repeated split sampling
Donhauser, J., Doménech-Pascual, A., Han, X., Jordaan, K., Ramond, J.-B., Frossard, A., Romaní, A.M., Priemé, A., 2024. Modelling soil prokaryotic traits across environments with the trait sequence database ampliconTraits and the R package MicEnvMod. Ecological Informatics 83, 102817. doi:10.1016/j.ecoinf.2024.102817