I would like to compare parameters across groups in a dataset which I have used multiple imputation (via aregImpute). First, several continuous parameters are such that I would like to use the non-parametric Kruskal-Wallis test, however I don't see this as an option using fit.mult.impute in rms. Second, I also am not finding an easy way to use fit.mult.impute for a Chi-square test for the categorical parameters. Perhaps I am overlooking something or approaching this incorrectly? Essentially what I have is below:
imputed_data <- aregImpute(~ PARAMS, n.impute=100, nk=4, data=df,type = "pmm", match = "weighted", burnin = 100)
I would like to then use fit.mult.impute to generate pooled estimates for both the Kruskal-Wallis and Chi-square tests across a grouping parameter in the data:
model_cont <- fit.mult.impute(continuous_parameter ~ group_parameter, fitter = ???, xtrans = imputed_data, data = df) model_cat <- fit.mult.impute(categorical_parameter ~ group_parameter, fitter = ???, xtrans = imputed_data, data = df)
Any thoughts on how to accomplish this? Thank you!