Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Big-O notation - correction

Author: Andrew Dados

Date: 12:18:12 05/09/01

Go up one level in this thread


On May 09, 2001 at 15:11:41, Eugene Nalimov wrote:

>Ok, here is simple example: let's consider sorting but limit the input by (a)
>limiting # of digits in each number (e.g. 1024 bits), and (b) limiting # of
>elements in the sequence we want to sort (e.g. max 10**40 elements). Yes,
>theoretical # of operations that are necessary to sort such a sequence is
>limited. But, nevertheless, you still can say that
>* insertion sort is O(N*N),
>* quick sort is O(N*ln(N)) on average, but O(N*N) in the worst case,
>* heap sort is O(N*ln(N)) in the worst case.
>So I don't see why the fact that something is finite changes something in the
>complexity tree search.
>
>Eugene

Wrong example. Instead consider sorting a set (2,42,15,103,37). Sort it; save a
solution.

Then what is a cost of sorting exact same set again?

There is infinite sets of numbers to sort; there is only one chess position to
solve....

-Andrew-



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.