[Haskell-cafe] Re: Purely funcional LU decomposition

[Michael D. Adams]

On Wed, Feb 4, 2009 at 2:15 AM, Neal Alexander <wqeqweuqy at hotmail.com> wrote:
> Array is no good man! Quad Tree matrices perform much nicer from what I've
> seen.

I've co-authored a paper on that topic ("Seven at one stroke: results
from a cache-oblivious paradigm for scalable matrix algorithms"
http://portal.acm.org/citation.cfm?id=1178597.1178604).  The current
implementation is C based (to get base-case performance) but we would
love to get that work ported to Haskell and a much more functional
setting.

Michael D. Adams