r - caret: Partition cases from multiply imputed datasets -
i ran mice
on dataset:
#run mice df <- mice::nhanes imp <- mice(df) #imputes data com <- complete(imp, "long", true) #creates data frame; raw data appended
now, want create training , testing datasets subsequent analyses. caret
can used split data, don't know concise way ensure same cases selected across imputations:
set.seed(1) splitindex <- createdatapartition(y = com$.id, p = 0.8, list = false) #rows considered independent; cases not selected consistently
any suggestions?
edit:
i have answered own question below.
Comments
Post a Comment