@ricardoV94 a follow up question, the mask approached work fine when the dataset its fixed
however when using a MutableData its not possible to use indexing
IndexError: only integers, slices (`:`), ellipsis (`...`), numpy.newaxis (`None`) and integer or boolean arrays are valid indices
I have some MutableData and some coords_mutable, how can I do a mutable mask ? is there a way to model this problem in order to re-train the model in other datasets ?