Big-O notation

We use big-Θ notation to asymptotically bound the growth of a running time to within constant factors above and below. Sometimes we want to bound from only above. For example, although the worst-case running time of binary search is $\Theta(\lg{n})$, it would be incorrect to say that binary search runs in $\Theta(\lg{n})$ time in all cases. What if we find the target value upon the first guess? Then it runs in $\Theta(1)$ time. The running time of binary search is never worse than $\Theta(\lg{n})$, but it's sometimes better. It would be convenient to have a form of asymptotic notation that means "the running time grows at most this much, but it could grow more slowly." We use "big-O" notation for just such occasions.

If a running time is $O(f(n))$ ...