Bill Allombert on Thu, 3 Apr 2003 22:42:54 +0200

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: polredabs() again

On Wed, Mar 12, 2003 at 03:09:12PM -0500, Igor Schein wrote:
> Hi,
> I know there're too many things going on at the same time right now,
> but I'll still gonna mention this example of polredabs() inefficiency:
> polredabs(x^8-97418273277417164826810*x^4-97418273278115084139045*x^2+97418273278115084139045);
> It takes extremely long time.  Basically, I'd like to see either a
> quick fix ( if possible ) or an acknowledgement of the issue in TODO
> file :).  I know we've discussed this with Karim in private emails,
> but it never made it into TODO.  

[I was expecting Karim to comment but...]

Well, the underlying algorithm has a high complexity, so it is not clear
it is a defficiency in PARI or even that it is fixable, if we insist
to have the 'absolutly smallest polynomial'.

Do you have experienced with other implementation (Kant OrderShort
function ?)