Proper way to “say” Big-O notation?

What is the proper way to convey an algorithm complexity in Big-O notation in speech?

Dose "the total number of operations is big oh of N log N" sounds strange?

What's generally accepted:

"It have "order log n" space complexity"?

"It is guaranteed to run in n log n time"?

-
"Ow", as in: my head hurts... – sdg Apr 20 '11 at 12:41
Big-"oh" <garbage> – Wildling Oct 8 '11 at 9:23

Personally I would just say "this algorithm is n log n". Anyone who is going to understand what that means will understand it from that phrasing; anyone who isn't, won't be helped by anything more verbose.

-

I say "order n" (or n squared etc).

If you are among computer scientists you may need to specify "big-O" or "omega" etc, as they are different, but most programmers simply refer to "order" meaning big-O. See here for more gory details: http://en.wikipedia.org/wiki/Big_O_notation

-

If you are speaking in a formal setting, such as making a presentation, it's good to say "This algorithm runs in time big-oh n log n".

If you are talking casually with another engineer (who also understands these concepts), you can be less formal and say "this runs in n log n time" or "this runs in order n log n time" or "this algorithm is n log n".

-

Just as f(n) is pronounced "f of n" and y(x) is "y of x", so O(n^2) is pronounced "o of n squared" or sometimes "big o of n squared."

-

I usually just say 'oh', as in:

Bubble sort is oh n squared

or:

Binary searches are oh log n

-

Every time i need to talk about Big O notation, lets say "Oh this algorithm Big O complexity is X", i try to explain it like that an algorithm complexity is at most X (where X could be n , n^a , a^n , log n, etc.), so the algorithm will behave in the worst case like the function X.

In most cases there is no need to specify Big O but there are 2 more methods to calculate complexity. One tries to look for a minimun bound, so the explanation now is "The algorithm behaves in the best case like X" , and finally there is a way to use stadistics to try to determine how does your algorithm work in most cases.

-