The article isn't concerned with estimates, but if you have large amounts of data, approximate numbers may be sufficient. I am a fan of HyperLogLog. It can be u

The article isn't concerned with estimates, but if you have large amounts of data, approximate numbers may be sufficient. I am a fan of HyperLogLog. It can be used as an online algorithm, so you don't have to keep all values memory. It could be useful if you want to do something like estimate the number of unique links posted to twitter in week. (if you could get that data) https://en.wikipedia.org/wiki/HyperLogLog https://en.wikipedia.org/wiki/HyperLogLog