Re: extreme elllseries() behavior

[Karim Belabas <Karim.Belabas@math.u-psud.fr>]

* Igor Schein [2003-08-20 10:18]:
> On Sat, Aug 02, 2003 at 04:00:01PM +0200, Karim BELABAS wrote:
> > On Fri, 1 Aug 2003, Igor Schein wrote:
> > > ? forstep(k=23,24,.1,print1(k" ");print(elllseries(e,k)))
> > > 23 0.9999999999999715783271077903
> > > 23.10000000000000000000000000 0.9999999999999752574949227059
> > > 23.20000000000000000000000000 0.9999999999999784603968424610
> > > 23.30000000000000000000000000 0.9999999999999812486647844467
> > > 23.40000000000000000000000000 0
> > > 23.50000000000000000000000000
> > >   ***   object too big, length can't fit in a codeword
> > 
> > And what is 'e' here ?
> 
> sorry, ellinit([0,0,1,0,10])
> 
> Using cut-off point !=1 allows to work around it, but the higher s is,
> the further A needs to be from 1, slowing things down considerably, as
> documented.  In the actual example, error, though too general, at least
> prevents the disaster, while value of 0 at s=23.4 is unacceptable IMO.

I believe the above is now fixed in CVS [ and elllseries should be able to
handle complex arguments ]

Cheers,

    Karim.
-- 
Karim Belabas                     Tel: (+33) (0)1 69 15 57 48
Dep. de Mathematiques, Bat. 425   Fax: (+33) (0)1 69 15 60 19
Universite Paris-Sud              http://www.math.u-psud.fr/~belabas/
F-91405 Orsay (France)            http://pari.math.u-bordeaux.fr/  [PARI/GP]