coderfender commented on PR #21817:
URL: https://github.com/apache/datafusion/pull/21817#issuecomment-4527454746
I added another check to not prevent multiple membership tests in low
density join use cases . Below are the new numbers :
```
group hashmap
roaring_bitmap
----- -------
--------------
hash_join_semi_anti/right_anti_d100_h10/1000000 3.33 5.3±0.05ms
? ?/sec 1.00 1587.1±9.07µs ? ?/sec
hash_join_semi_anti/right_anti_d100_h100/1000000 1.86 3.3±0.02ms
? ?/sec 1.00 1789.2±6.90µs ? ?/sec
hash_join_semi_anti/right_anti_d10_h10/1000000 4.36 11.3±0.18ms
? ?/sec 1.00 2.6±0.01ms ? ?/sec
hash_join_semi_anti/right_anti_d10_h100/1000000 1.63 5.3±0.05ms
? ?/sec 1.00 3.3±0.03ms ? ?/sec
hash_join_semi_anti/right_anti_d50_h10/1000000 2.70 5.3±0.08ms
? ?/sec 1.00 1979.4±13.13µs ? ?/sec
hash_join_semi_anti/right_anti_d50_h100/1000000 1.56 3.3±0.02ms
? ?/sec 1.00 2.1±0.01ms ? ?/sec
hash_join_semi_anti/right_semi_d100_h10/1000000 1.00 1496.0±27.48µs
? ?/sec 1.05 1572.4±8.69µs ? ?/sec
hash_join_semi_anti/right_semi_d100_h100/1000000 4.48 8.1±0.12ms
? ?/sec 1.00 1797.5±20.95µs ? ?/sec
hash_join_semi_anti/right_semi_d10_h10/1000000 2.85 7.4±0.15ms
? ?/sec 1.00 2.6±0.01ms ? ?/sec
hash_join_semi_anti/right_semi_d10_h100/1000000 3.12 10.1±0.21ms
? ?/sec 1.00 3.2±0.01ms ? ?/sec
hash_join_semi_anti/right_semi_d50_h10/1000000 1.00 1488.1±10.98µs
? ?/sec 1.31 1954.4±7.02µs ? ?/sec
hash_join_semi_anti/right_semi_d50_h100/1000000 3.76 8.1±0.07ms
? ?/sec 1.00 2.1±0.01ms ? ?/sec
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
