I recently learned about HyperLogLog, a datastructure for fast approximate distinct counts. I made a little blog post about implementing it:
I recently learned about HyperLogLog, a datastructure for fast approximate distinct counts. I made a little blog post about implementing it: