Re: [eigen] first benchmark of large and sparse matrix

[ Thread Index | Date Index | More Archives ]

On Monday 23 June 2008 07:05:30 Benoît Jacob wrote:
> this is of high interest for Krita and for Step and probably has
> countless interesting applications

I forgot to mention that support for sparse matrices should be of high 
interest to OpenBabel (chemistry lib, also considering eigen2): if I remember 
well they sometimes apply solving algorithms to the adjacency matrix of a 
molecule, so for a molecule with N atoms the matrix is NxN and has 2N nonzero 
entries. As far as I remember they used to apply dense algorithms, so N^3 
complexity which is only doable for small N (like 10^2) while some molecules 
they deal with have up to 10^5 atoms.

This example suggests that adjacency matrices of large graphs are a countless 
source of applications for sparse matrices...



Attachment: signature.asc
Description: This is a digitally signed message part.

Mail converted by MHonArc 2.6.19+