On Tue, 6 Nov 2007 [EMAIL PROTECTED] wrote:

> However, this is still an O(log n) algorithm, because that's the
> complexity of raising-to-the-power-of.  And it's slower than the
> simpler integer-only algorithms.

You mean computing the matrix power of

/1 1\
\0 1/

?
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe

Reply via email to