doris
doris copied to clipboard
[improve](percentile_approx) use pdqsort to improve percentile_approx function performance
Proposed changes
could faster about 1s on table rows of count 150000000