Hi all!   I have a data set with 2million obs and I want to make 7 groups out of it (6 groups with 300,000 obs and one control group with 200,000 obs). It has to be a random selection, however it has to follow this rule:    There is 'country' input in the data and the random groups must have the same country percentages as the big data itself, i.e. if United States is 40% of the data, then all of these subgroups must have 40% United States in the country column.    Any idea is highly appreciated!   Thanks 🙂 ... View more