on Thursday 20 November 2003 23:51 WJ wrote:
>There is also share/numeric/qq.lisp, which is described in the manual.
>The CVS version doesn't work, but it has been fixed
>
>https://sourceforge.net/tracker/?func=detail&atid=104933&aid=842863&group_id
>=4933
>
>and it is quite a bit faster than romberg on certain problems (try
>qq.dem). I will commit it soon.
So, qq is revived. Great! I remeber it was ok in 5.6 but after some moment it
became broken. I tested files you pointed out and find that new qq works fine
on my integrands.They are mixed 2'd and 3'd kind elliptic integrals. It is
roughly 5 times faster than romberg.
rgds,
V