We compute  entanglement spectrum of Hubbard model with real terms

using ALPS easily. When an imaginary term as follow

<BONDOPERATOR name="DSO_coupling" source="x" target="y">
  I*cdag_up(x)*c_up(y)-I*cdag_down(x)*c_down(y)-I*cdag_up(y)*c_up(x)+I*cdag_down(y)*c_down(x)
</BONDOPERATOR>

is added to Hamiltonian, the entanglement spectrum is very sensitive to the size of lattice and MAXSTATES.

Can ALPS compute the entanglement spectrum of any Hermitian Hamiltonian with non-real terms?

Thanks for your attention