Skip to content

Comment on Quadsort: a stable non-recursive merge sortparent

Comments

It depends on your model of computation.

O(n log n) is only the frontier for comparison based sorts that know nothing about the distribution of inputs.

If your sorting algorithm is allowed to do anything else on your data, like hashing or looking at bits or arithmetic, different lower bounds might apply.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.