Somewhere in the E.U., le 02/06/2005 Bonjour In article <008e01c566c1$e31850b0$2b792518@heronnest>, christopher diggins <cdiggins@videotron.ca> wrote:
I have just finished writing a fixed dimensionality matrix template (kmatrix<T, Rows, Cols>) which appears to significantly outperform the ublas::matrix. Is there any interest? Below are the results of the benchmarks using Visual C++ 7.1 on an Intel Celeron 1.6 GHZ:
[SNIP]
Christopher Diggins http://www.cdiggins.com
I believe we badly need some de jure standard for C++ incarnations of common mathematical concepts, in particular w.r.t. algebra and (euclidian) geometry. "Small" matrices are in that group. The fact that they would be fast would be icing on the cake, not the cake itself, as it were. What is most needed IMHO is a clear and rich set of links between the various elements of that collection (complex <-> plane (vector) rotation; rotation <-> geometric elements of the rotation (when meaningful); 2D, 3D, 4D points;...). Everybody and his (or her) favorite pet has a homegrown version of that kind of library, more or less naïve and more or less efficient, but what we need is really something which we could present as a standard, and speed (alone) is not sufficient for that to happen. Expressivity and richness of functionality would be better bets. Care must also be taken not to sever the bridges with more general packages such as ublas, of course. At any rate, many such proposals have been floated here but, AFAIK, none has ever been developed to the point of requesting a review, which I feel is really a shame. Merci Hubert Holin