3 ms·
> We won't have false positives (like with a bloom filter) -- we'll have a degree of error (i.e. the actual count may be 1000, but the HLL reports 1004). Isn’t
by mrhands556 4y ago
> We won't have false positives (like with a bloom filter) -- we'll have a degree of error (i.e. the actual count may be 1000, but the HLL reports 1004).
Isn’t what the article is describing a false positive? I expected an undercount.