IsInverseSemigroup is inconsistent for some bipartition semigroups

Issue #144 resolved
Attila Egri-Nagy
created an issue
gap> S := Semigroup([ Bipartition( [ [ 1, 2 ], [ -1 ], [ -2 ] ] ), 
>   Bipartition( [ [ 1, -1 ], [ 2 ], [ -2 ] ] ), 
>   Bipartition( [ [ 1 ], [ 2, -1 ], [ -2 ] ] ), 
>   Bipartition( [ [ 1, -2 ], [ 2 ], [ -1 ] ] ), 
>   Bipartition( [ [ 1 ], [ 2, -2 ], [ -1 ] ] ), 
>   Bipartition( [ [ 1 ], [ 2 ], [ -1 ], [ -2 ] ] ) ]);;
gap> IsInverseSemigroup(S);
false

while

gap> S := Semigroup([ Bipartition( [ [ 1, 2 ], [ -1 ], [ -2 ] ] ), 
>   Bipartition( [ [ 1, -1 ], [ 2 ], [ -2 ] ] ), 
>   Bipartition( [ [ 1 ], [ 2, -1 ], [ -2 ] ] ), 
>   Bipartition( [ [ 1, -2 ], [ 2 ], [ -1 ] ] ), 
>   Bipartition( [ [ 1 ], [ 2, -2 ], [ -1 ] ] ), 
>   Bipartition( [ [ 1 ], [ 2 ], [ -1 ], [ -2 ] ] ) ]);;
gap> NrDClasses(S);;
gap> IsInverseSemigroup(S);
true

It seems that calculating the number of D-classes changes the answer.

Comments (4)

  1. Log in to comment