an-cabal/an-rope

"Fuzzy" balance heuristic

Opened this issue · 0 comments

hawkw commented

Currently, we always rebalance a Rope if it is out of balance. A potential way to improve rebalance performance might be changing the is_balanced() method to tolerate cases where the Rope is only slightly out of balance, given that if balance gets worse, it will be rebalanced again anyway. This way, we don't rebalance() the rope quite as much, but it still stays reasonably balanced.