How can I tell Spark that it can now discard / free / garbage a Dataset, when I know that I won't use it anymore?

Question

From dataset A and B, I\'ve done a brand new dataset C.

Let\'s say that I know that I won\'t use A any