The documentation on subsets is a bit sparse. Is there a way of creating more complex strategies for subsetting?
i.e. interactions often occur between tables: products, orders, customers, a percentage of products might leave a lot of empty orders, or a lot of products not bought.
What if I wanted:
a subset of users based on a random sample
a subset of those users orders, maybe based on a specific time period (the last three months)
a subset of products—maybe all the ones in the orders above, plus a few random extra ones
The documentation on subsets is a bit sparse. Is there a way of creating more complex strategies for subsetting?
i.e. interactions often occur between tables: products, orders, customers, a percentage of products might leave a lot of empty orders, or a lot of products not bought.
What if I wanted: