Closed dbosak01 closed 3 years ago
The worst problem appears when using a dyplyr group_by function. If you use the by parameter on the datastep instead of the group_by function, it goes fast.
Huge increase gained by converting from grouped_tbl to data.frame. Still getting a freaking problem on check that doesn't show up on test. Will open a separate issue.
Datastep is still too slow. Need another 3X improvement.