world-federation-of-advertisers / cardinality_estimation_evaluation_framework

Evaluation framework and methods for estimating cardinalities of groups of sets
Apache License 2.0
22 stars 9 forks source link

update HLL++ to use sparse mode #96

Closed jgoodknight closed 4 years ago

jgoodknight commented 4 years ago

HLL++ doesn't do well for small cardinalities so we are updating our implementation to use a version of sparse mode from the HLL++ paper, that doesn't care so much about memory space but is much simpler to implement.

Note this is not tested.

huangxichen1 commented 4 years ago

Please add tests, especially for the merge.

Added tests.

Please add tests, especially for the merge.

I added the tests. PTAL. Thanks!