nspin out of bounds

Submitted by kns43b on Thu, 03/23/2017 - 07:40

Forums 

User questions

I'm trying to run sigma.cplx.x, and I'm getting the following error:

From proc 0: ERROR: nspin out of bounds

Inspection of WFN_inner with wfn_rho_vxc_info.x shows a number of spins equal to 1. The epsilon calculations (split over qpoints) seem to have run without complaint, and both WFNq.h5 and WFN.h5 also show an nspin of 1. The sigma.out file reads as follows:

...
Using the Generalized Plasmon Pole model

We are communicating via MPI

Computing CH as a partial sum over empty bands with the static remainder

Running with semiconductor screening

We are using a truncated Coulomb interaction: Cell Slab

Cutoff for Monte-Carlo average of Coulomb potential: 1.000E-012 Ry

Reading exchange-correlation matrix elements from vxc.dat file

Treating W within the Hybertsen-Louie Generalized Plasmon Pole model

Memory available: 3038.2 MB per PE

Read from eps0mat.h5:
- Number of G-vectors in the charge-density grid: 374539
- Max. number of G-vectors for dielectric matrices: 2371
- Cutoff of the dielectric matrix (Ry): 34.000
- Number of imaginary frequencies: 0
- Total number of frequencies: 1
- Frequency dependence: 0
- Number of q-points: 1
- Q grid: 8 8 1

Read from epsmat.h5:
- Number of G-vectors in the charge-density grid: 374539
- Max. number of G-vectors for dielectric matrices: 2379
- Cutoff of the dielectric matrix: 34.000
- Number of imaginary frequencies: 0
- Total number of frequencies: 1
- Frequency dependence: 0
- Number of q-points: 63
- Q grid: 8 8 1

Reading header of WFN_inner
Highest occupied band (unshifted grid) = 14
Valence max (unshifted grid) = -1.904259 eV
Conduction min (unshifted grid) = 1.030961 eV
Middle energy (unshifted grid) = -.436649 eV
Fermi energy (unshifted grid) = -.436649 eV

Calculation parameters:
- Cutoff of the bare Coulomb interaction (Ry): 250.00
- Cutoff of the screened Coulomb interaction (Ry): 34.00
- Number of G-vectors up to the bare int. cutoff: 46699
- Number of G-vectors up to the screened int. cutoff: 2371
- Total number of bands in the calculation: 899
- Number of fully occupied valence bands: 14
- Number of partially occ. conduction bands: 0

Memory required for execution: 265.1 MB per PE
Memory required for vcoul: 114.4 MB per PE

Number of electrons per unit cell (from ifmax) = 28.000000
Number of electrons per unit cell (from occupations) = 28.000000
Plasma Frequency = 1.416197 Ry

Q-grid symmetries are being used.
------

Are there extra considerations I should take into account when working with 2D materials, or are there any places in particular I should look to see what may be causing this issue?

Submitted by babarker on Mon, 03/27/2017 - 12:33

Hello,

Did you also check the charge density file, RHO, with wfn_rho_vxc_info.x?

Best,
Brad

Submitted by kns43b on Tue, 04/04/2017 - 13:23

Hi,

I did not check RHO. After doing so though, it seems that indeed that's where the error was at. Thanks for the help.

Submitted by babarker on Wed, 04/05/2017 - 18:04

Hello,

Good to hear that we figured it out. If you created a 'VXC' binary file, you might also want to check and re-generate that, too.

Best,
Brad