An indication of the relative performance of an algorithm. An algorithm or the method that implements it is said to be O(1) if its execution is a constant, O(log n) if its execution is proportional to the natural log of n, or O(n) if its execution is proportional to n. O notation is frequently used as a way to compare the relative performance of operations on collections, in which case n is usually the number of elements in the collection.