Karim BELABAS on Mon, 15 Sep 2003 18:20:19 +0200 (MEST)

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

Re: qfbpowraw

On Mon, 15 Sep 2003, Bill Allombert wrote:

> On Sun, Sep 14, 2003 at 04:39:07PM +0200, Karim BELABAS wrote:
> > On Sun, 14 Sep 2003, Bill Allombert wrote:
> > > It seems that PARI * and ^ operators allow operation on quadratic
> > > forms but reduce them. I would personnaly prefer it does not reduce
> > > them.
> >
> > Coefficient explosion would occur for ^.
> ...only if you plan to reduce the result.

I don't understand this.

Currently, if you want no reduction, you use xxxraw routines, whereas the
natural operators (*, ^) do reduce.

With your proposed change (which I don't like), the natural operators would
not reduce and you would need to apply reduction yourself if you want it.
Except that for ^ it's not possible to do this at the proper place [ you can
only reduce at the end, not after each composition ].

> Please note that the == operator is not consistent with this
> policy, nor is ^1.

It is, assuming all inputs are reduced [ at least for t_QFI ]

> > qfb[comp|pow]raw would be dumped then ?
> Why ?

Because they would serve no purpose if * and ^ were doing the same. But we
would need to introduce a qfbpowred function.

> qfbnucomp/qfbnupow are still available.

These are only applicable to t_QFI, not t_QFR. These 2 do reduce their output
of course.

Note that t_QFI and t_QFR were only introduced for quick GP computations.
They are not used at all anywhere in the library. Even for quadclassunit()
routines, we use private more efficient versions. I always found these two
types were a waste of ressources...

I don't see the point of breaking backward compatibility in this way.


Karim Belabas                     Tel: (+33) (0)1 69 15 57 48
Dép. de Mathématiques, Bât. 425   Fax: (+33) (0)1 69 15 60 19
Université Paris-Sud              http://www.math.u-psud.fr/~belabas/
F-91405 Orsay (France)            http://www.parigp-home.de/  [PARI/GP]