-
Notifications
You must be signed in to change notification settings - Fork 19
Open
Labels
bugAny bug should have this label, even if it also has a more generic labelAny bug should have this label, even if it also has a more generic labelbug: wrong answerA computation returns a (mathematically or otherwise) invalid resultA computation returns a (mathematically or otherwise) invalid resultmandarinsThis issue is likely to be resolved by the addition of mandarinsThis issue is likely to be resolved by the addition of mandarins
Description
The following loop quite quickly leads to an incorrect size:
gap> testFDPM := function(deg, field, alternating)
> local g, gens, mgens, m, cf, dims, max, pos, x, h, result;
> if alternating then
> g := AlternatingGroup(deg);
> else
> g := SymmetricGroup(deg);
> fi;
> gens := List([1..5],x->PseudoRandom(g));
> mgens := List(gens,x->PermutationMat(x,deg,field));
> m := GModuleByMats(mgens,field);
> cf := MTX.CompositionFactors(m);
> dims := List(cf,x->x.dimension);
> max := Maximum(dims);
> pos := Position(dims,max);
> x := PseudoRandom(GL(max,field));
> mgens := List(cf[pos].generators,y->y^x);
> h := Group(mgens);
> return Size(RecogniseGroup(h)) = Size(g);
> #return RECOG.TestGroup(h,false,Size(g));
> end;;
gap> bool := false;
gap> while testFDPM(10, GF(7), bool) do
> bool := not bool;
> od;
This probably will be solved once #265 is merged in. But, that would only hide the error underlying this issue. Somewhere either the immediate verification is not strong enough or there is a bug.
Metadata
Metadata
Assignees
Labels
bugAny bug should have this label, even if it also has a more generic labelAny bug should have this label, even if it also has a more generic labelbug: wrong answerA computation returns a (mathematically or otherwise) invalid resultA computation returns a (mathematically or otherwise) invalid resultmandarinsThis issue is likely to be resolved by the addition of mandarinsThis issue is likely to be resolved by the addition of mandarins