OHE calc.: The absolute values of the conductivity depend on the used version

[AKB-OU]

Dear developers and users,

Hi.

In the latest release version 2.6.2, I encountered a problem in the OHE calculation that sigma_xx (and thus rho_xx) become negative at high B*tau values. This problem does not arise in the version 2.5.1, which is tested with exactly the same inputs.

I noticed that the lattice information seems to be wrongly recognized in 2.6.2: In 2.5.1, my tetragonal system is properly recognized like

lattice information (Angstrom) a b c alpha beta gamma 4.349498 4.349498 10.559891 90.000000 90.000000 90.000000 Three Lattice vectors of the unfolded cell: 4.349498 0.000000 0.000000 0.000000 4.349498 0.000000 0.000000 0.000000 10.559891

But in 2.6.2, the angles alpha, beta, and gamma are strange like

lattice information (Angstrom) a b c alpha beta gamma 4.349500 4.349500 10.559895 47.625949 47.625949 47.625949 Three Lattice vectors of the unfolded cell: 4.349500 0.000000 0.000000 0.000000 4.349500 0.000000 0.000000 0.000000 10.559895

In addition, "Atom position and magnetic moment of Original lattice" in WT.out depends on the used version.

I attached my TB hamiltonian and wt.in for your reference. hr_and_input.zip

Best regards, AKB-OU

[AKB-OU]

Dear all,

Happy new year.

I have not been able to solve the problem yet. The absolute values of the conductivity depend on the used version.

I attach a comparison of the result obtained by v2.5.1 and latest versions. LAS_comp.pdf

By the way, the above phenomena can be seen in Cu tutorial as attached below. Cu_theta0_comp.pdf

Best, AKB

[Dongsheng-Wen]

Looks like a factor of two is there between the versions. I was able to reproduce the results of Cu between v2.7.0 and v2.5.1. The plot matches after x2 for the results calculated by v2.7.0.

[AKB-OU]

Dear Dongsheng-Wen,

Thank you for the information. In my opinion, this problem has remained unsettled in the versions >= 2.6.

Best, AKB