I am working with data containing observations for patients undergoing a surgical procedure that is commonly (but not always) performed by access in both femoral/groin arteries. These arteries may be accessed via an open surgical incision or via a minimally-invasive percutaneous approach. My dataset has a variable for the manner in which each side was or was not accessed (e.g. 0 == not accessed, 1 == open surgical, 2 == minimally-invasive). In order to examine these access sites and their methods of access, I wish to generate duplicates for those patients who had both sides accessed, with each copy containing the entry for one access side and the other blanked.