This add-in allows the user to split a dataset into train/validate/test partitions. It includes options for rebalancing the proportions of the output data set's strata variable levels in relation to a focal group. This feature is useful, for example, in oversampling an event that is rare in the original data.
Instructions for using the add-in are attached.
Updated 3/23/2016: Includes additional balancing options.
Updated 9/1/2016: Bug fixes (related to an error when running the add-in)