[IPOL discuss] about performance comparisons in IPOL articles

[Nicolas Limare]

> I agree with this.
> However, I think that a fair comparison of two algorithms shouldn't
> be only based on the execution time, but on the complexity analysis.

... but a complexity analysis is not always possible or practical,
for example when some informaton is missing or when we only have a
compiled binary implementation. And big-O notation doesn't tell you,
practically, which alfgorithm to choose. If algo A is in O(n2) and algo
B in O(n log n) but cost(A) is only higner than cost(B) for n>10^100,
then for all real-life input data you are better with algo A.

I'd say that both metrics (exec time and complexity) provide different
and complementary info on an algorithm.

-- 
Nicolas LIMARE - CMLA - ENS Cachan         http://limare.perso.math.cnrs.fr/
IPOL journal                                             http://www.ipol.im/
-> image processing, reproducible research, open science
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <https://tools.ipol.im/mailman/archive/discuss/attachments/20130612/cbd5eabd/attachment.pgp>