Dear Javier,
"JRG" == Javier Ros Ganuza <jros@unavarra.es> writes: JRG> I'll will install CVS version to follow.
I think it should go smooth. JRG> Mean while, I see that geometric products of type e~1*e~1 are JRG> being simplified to ONE, that I suppose that can be removed JRG> using JRG> ex remove_dirac_ONE(const ex & e) That function is intended only for an expression e which has a form (a Clifford scalar) * ONE. You could not remove ONE from the expression 2*ONE+3*e~1, since it will become an invalid GiNaC expression (a commuting object cannot be added to a non-commuting one). JRG> Also I see in JRG> a0+(e~0*e~1)*a3+e~1*a2+a1*e~0 JRG> that e~1*e~0 expresions are outputed like (e~0*e~1). In this way GiNaC highlights that (e~0*e~1) is a non-commutative product times some other commuting terms. JRG> If I want to trigger specific simplifications for the Cl2 JRG> geometric algebra, like JRG> (e~1*e~0)=-(e~0*e~1) GiNaC just put in expression in a "canonical form" which means some "canonical order". If you simply declare e00 and e01 in the reverse order GiNaC would output (e~1*e~0) as -(e~0*e~1). JRG> e~0*(e~0*e~1)=e~1 JRG> e~0*e~1*e~0*e~1=-1 All that should be done by canonicalize_clifford() method. Just remember that it should be called manually whenever you need it. Best wishes, Vladimir -- Vladimir V. Kisil email: kisilv@maths.leeds.ac.uk -- www: http://maths.leeds.ac.uk/~kisilv/