Dear ALPS users,
I would like to build a non-symmetric Hamiltonian. To do so I started playing with definitions, defining


<SITEOPERATOR name="Sx" site="x">
  1/2*(Splus(x)+Sminus(x))
</SITEOPERATOR>

<SITEOPERATOR name="Syy" site="x">
  1/2*(Sminus(x)-Splus(x))
</SITEOPERATOR>


and the two Hamiltonians

<HAMILTONIAN name="H1">
  <PARAMETER name="Kappa" default="0"/>
  <BASIS ref="spin"/>
  <SITETERM site="i">
    Sz(i)+Kappa*Syy(i)
  </SITETERM>
</HAMILTONIAN>


<HAMILTONIAN name="H2">
  <PARAMETER name="Kappa" default="0"/>
  <BASIS ref="spin"/>
  <SITETERM site="i">
    Sz(i)+Kappa*Sx(i)
  </SITETERM>
</HAMILTONIAN>


Diagonalizing exactly these Hamiltonians for few sites of an open chain I expect different results, because there's no unitary transformation that connects Sx with Syy (there would be between Sx and Sy, but Syy=-i*Sy).

Strangely enough I get the same eigenvalues for both Hamiltonians.
Do you know what is the problem here? Am I allowed at all to define antisymmetric operators in ALPS?
Thank you in advance for your help.

Kindest regards.

--
Dr Emanuele Levi

emanuele.levi@gmail.com