Re: polredabs()

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

[Igor:]
> ? t2norm(x^4-x-1)
> %13 = 4.276475609680439616479500351
> ? t2norm(polredabs(x^4-x-1))
> %14 = 4.345048321046443827557722074
> 
> I think though it is a 2.0.12-introduced bug, because
> in 2.0.11 polredabs(x^4-x-1) would return itself, rather than
> x^4-x^3-1, like 2.0.12 does.  

Right. A typo crept in while performing the final change of basis [inverting
a permutation, which is almost always = identity]. Here's a patch.

Karim.

*** src/basemath/bibli1.c.orig  Fri Nov  6 16:08:07 1998
--- src/basemath/bibli1.c       Tue Dec  8 14:23:37 1998
***************
*** 3106,3112 ****
    perm = sindexsort(vnorm);
    sperm = cgetg(n,t_MAT);
    uperm = cgetg(n,t_MAT);
!   for (i=1; i<n; i++) { uperm[perm[i]] = u[n-i]; sperm[n-i] = s[perm[i]]; }

    if (DEBUGLEVEL>2) {fprintferr("\nEntree dans smallvectors\n"); flusherr();}
    if (check) basis = init_chk(chk_nf,uperm);
--- 3105,3111 ----
    perm = sindexsort(vnorm);
    sperm = cgetg(n,t_MAT);
    uperm = cgetg(n,t_MAT);
!   for (i=1; i<n; i++) { uperm[n-i] = u[perm[i]]; sperm[n-i] = s[perm[i]]; }

    if (DEBUGLEVEL>2) {fprintferr("\nEntree dans smallvectors\n"); flusherr();}
    if (check) basis = init_chk(chk_nf,uperm);

--
Karim Belabas                    email: Karim.Belabas@math.u-psud.fr
Dep. de Mathematiques, Bat. 425
Universite Paris-Sud             Tel: (00 33) 1 69 15 57 48
F-91405 Orsay (France)           Fax: (00 33) 1 69 15 60 19
--
PARI/GP Home Page: http://pari.home.ml.org