The short version is, there doesn't appear to be any change between the times for non-empty masks.
Additionally, this branch contains unit tests verifying the new behavior. @tpietzsch please let me know if you have any questions or if you'd like any changes.
The first commit contains the benchmark tests:
Output Before Change:
Output After Change:
The short version is, there doesn't appear to be any change between the times for non-empty masks.
Additionally, this branch contains unit tests verifying the new behavior. @tpietzsch please let me know if you have any questions or if you'd like any changes.
Closes #31