Testing latest pari + WASM + node.js... and it works?! Wow.
License: GPL3
ubuntu2004
Function: alggroupcenter
Section: algebras
C-Name: alggroupcenter
Prototype: GDGD&
Help: alggroupcenter(gal,{p=0},{&cc}): constructs the center of the group
algebra of gal over Q (resp. Fp), and sets cc to the conjugacy classes of gal.
Doc: initializes the center~$Z(K[G])$ of the group algebra~$K[G]$ over~$K=\Q$
($p = 0$ or omitted) or~$\F_p$ where~$G$ is the underlying group of the
\kbd{galoisinit} structure~\var{gal}. The input~\var{gal} is also allowed to
be a \typ{VEC} of permutations that is closed under products.
Sets~\var{cc} to a \typ{VEC}~$[\var{elts},\var{conjclass},\var{rep},\var{flag}]$
where~\var{elts} is a sorted \typ{VEC} containing the list of elements
of~$G$, \var{conjclass} is a \typ{VECSMALL} of the same length as~\var{elts}
containing the index of the conjugacy class of the corresponding element (an
integer between $1$ and the number of conjugacy classes), and~\var{rep} is a
\typ{VECSMALL} of length the number of conjugacy classes giving for each
conjugacy class the index in~\var{elts} of a representative of this conjugacy
class. Finally \var{flag} is $1$ if and only if the permutation
representation of $G$ is transitive, in which case the $i$-th element
of \var{elts} is characterized by $g[1] = i$; this is always the case
when \var{gal} is a \kbd{galoisinit} structure. The basis of~$Z(K[G])$ as
output consists of the indicator functions of the conjugacy classes in the
ordering given by~\var{cc}. Example:
\bprog
? K = nfsplitting(x^4+x+1);
? gal = galoisinit(K); \\ S4
? al = alggroupcenter(gal,,&cc);
? algiscommutative(al)
%4 = 1
? #cc[3] \\ number of conjugacy classes of S4
%5 = 5
? gal = [Vecsmall([1,2,3]),Vecsmall([1,3,2])]; \\ C2
? al = alggroupcenter(gal,,&cc);
? cc[3]
%8 = Vecsmall([1, 2])
? cc[4]
%9 = 0
@eprog