Reading psml pseudodojo potentials with spin-orbit

External optional components of Abinit (BigDFT, Wannier90...) and codes using/providing data from/to Abinit (AtomPAW, ONCVPSP, EXC, DP, Yambo...)

Moderators: ebousquet, bxu

Locked
tsirkin
Posts: 1
Joined: Mon Jan 20, 2020 7:59 pm

Reading psml pseudodojo potentials with spin-orbit

Post by tsirkin » Wed Jan 29, 2020 5:44 pm

Dear All,

I am trying to use the pseudodojo potentials (NC FR ONCVPSP v0.4) in psml format, but I get errors.
In particular, I am trying to calculate bismuth, but I tried a few other elements, and still get errors.

with the PBE:

Code: Select all

--- !ERROR
src_file: m_pspini.F90
src_line: 1050
mpi_rank: 0
message: |
    Pseudopotential file cannot give spin-orbit characteristics,
    while pspso(itypat)= 2.
    Action: check your pseudopotential and input files for consistency.
...
with PBEsol the error is different:

Code: Select all

--- !BUG
src_file: m_psxml2ab.F90
src_line: 205
mpi_rank: 0
message: |
    Your psml potential should have either scalar- or non- relativistic projectors
...

When I run siesta with the same psml potentials, the calculation runs without errors, but I get results which are too much different then I get by other codes (ABINIt with psp8 or VASP)

When doing scalar-relativistic calculation (with NC SR pseudos), everything works fine (both in abinit and siesta) and the results agree well.

Any I deas, if it is a bug in Abinit or in the pseudodojo database, or in libpsml?

Thank you,
Stepan Tsirkin.

Below are the details of my compilation:

I am using ABINIT version 8.10.3
configured with --with-dft-flavor=libxc
--with-trio-flavor=netcdf+psml
fallbacks to enable => libxc netcdf
libxmlf90 version 1.5.4
libpsml version 1.1.8


The output of abinit -b is :

Code: Select all

> /app/theorie/abinit-8.10.3/bin/abinit -b
 DATA TYPE INFORMATION: 
 REAL:      Data type name: REAL(DP) 
            Kind value:      8
            Precision:      15
            Smallest nonnegligible quantity relative to 1:  0.22204460E-15
            Smallest positive number:                       0.22250739-307
            Largest representable number:                   0.17976931+309
 INTEGER:   Data type name: INTEGER(default) 
            Kind value: 4
            Bit size:   32
            Largest representable number: 2147483647
 LOGICAL:   Data type name: LOGICAL 
            Kind value: 4
 CHARACTER: Data type name: CHARACTER             Kind value: 1
  ==== Using MPI-2 specifications ==== 
  MPI-IO support is ON
  xmpi_tag_ub ................   2147483647
  xmpi_bsize_ch ..............            1
  xmpi_bsize_int .............            4
  xmpi_bsize_sp ..............            4
  xmpi_bsize_dp ..............            8
  xmpi_bsize_spc .............            8
  xmpi_bsize_dpc .............           16
  xmpio_bsize_frm ............            4
  xmpi_address_kind ..........            8
  xmpi_offset_kind ...........            8
  MPI_WTICK ..................    1.0000000000000001E-009

 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 CPP options activated during the build:

                    CC_GNU                   CXX_GNU                    FC_GNU
 
 HAVE_FC_ALLOCATABLE_DT...             HAVE_FC_ASYNC         HAVE_FC_BACKTRACE
 
  HAVE_FC_COMMAND_ARGUMENT      HAVE_FC_COMMAND_LINE        HAVE_FC_CONTIGUOUS
 
           HAVE_FC_CPUTIME              HAVE_FC_EXIT             HAVE_FC_FLUSH
 
             HAVE_FC_GAMMA            HAVE_FC_GETENV   HAVE_FC_IEEE_EXCEPTIONS
 
          HAVE_FC_INT_QUAD             HAVE_FC_IOMSG     HAVE_FC_ISO_C_BINDING
 
  HAVE_FC_ISO_FORTRAN_2008        HAVE_FC_LONG_LINES        HAVE_FC_MOVE_ALLOC
 
           HAVE_FC_PRIVATE         HAVE_FC_PROTECTED         HAVE_FC_STREAM_IO
 
            HAVE_FC_SYSTEM          HAVE_FORTRAN2003        HAVE_LIBPAW_ABINIT
 
      HAVE_LIBTETRA_ABINIT                HAVE_LIBXC               HAVE_LINALG
 
        HAVE_LINALG_SERIAL                  HAVE_MPI                 HAVE_MPI2
 
       HAVE_MPI_IALLREDUCE        HAVE_MPI_IALLTOALL       HAVE_MPI_IALLTOALLV
 
        HAVE_MPI_INTEGER16               HAVE_MPI_IO HAVE_MPI_TYPE_CREATE_S...
 
               HAVE_NETCDF                HAVE_NUMPY             HAVE_OS_LINUX
 
                 HAVE_PSML         HAVE_TIMER_ABINIT              USE_MACROAVE
 
 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++


 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 === Build Information === 
  Version       : 8.10.3
  Build target  : x86_64_linux_gnu7.5
  Build date    : 20200120

 === Compiler Suite === 
  C compiler       : gnu7.5
  C++ compiler     : gnu7.5
  Fortran compiler : gnu7.5
  CFLAGS           :  -g -O2 -mtune=native -march=native
  CXXFLAGS         :  -g -O2 -mtune=native -march=native
  FCFLAGS          :  -g -ffree-line-length-none
  FC_LDFLAGS       : 

 === Optimizations === 
  Debug level        : basic
  Optimization level : standard
  Architecture       : amd_opteron

 === Multicore === 
  Parallel build : yes
  Parallel I/O   : auto
  openMP support : no
  GPU support    : no

 === Connectors / Fallbacks === 
  Connectors on : yes
  Fallbacks on  : yes
  DFT flavor    : libxc-fallback
  FFT flavor    : none
  LINALG flavor : netlib
  MATH flavor   : none
  TIMER flavor  : abinit
  TRIO flavor   : netcdf-fallback+psml

 === Experimental features === 
  Bindings            : @enable_bindings@
  Exports             : no
  GW double-precision : no

 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++


 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 Default optimizations:
   -O2 -mtune=native -march=native


 Optimizations for 43_ptgroups:
   -O0


 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++


miro_ilias
Posts: 10
Joined: Thu Jul 19, 2018 8:18 pm

Re: Reading psml pseudodojo potentials with spin-orbit

Post by miro_ilias » Sat Jul 11, 2020 10:33 pm

Hello, has this issue been fixed ? I am also interested in full relativistic calculations...

ebousquet
Posts: 469
Joined: Tue Apr 19, 2011 11:13 am
Location: University of Liege, Belgium

Re: Reading psml pseudodojo potentials with spin-orbit

Post by ebousquet » Mon Jul 13, 2020 11:57 am

Dear Miro,
Sounds like it is a problem in the psml files, which do not include the SOC. Could you make a test (possibly with the v9.x version) and tell me if it is still problematic. If so I can regenerate the ones you need with SOC in the mean time we update the whole pseudodojo table.
Best wishes,
Eric

dbennett1994
Posts: 7
Joined: Sat Mar 21, 2020 10:21 am

Re: Reading psml pseudodojo potentials with spin-orbit

Post by dbennett1994 » Sat Nov 14, 2020 8:17 pm

Hi, I just tried to run a SOC calculation using fully relativistic psml pseudos from pseudodojo and got the same error. I tried with the corresponding psp8 pseudos and they seem to work. I'm using abinit 8.10.3 if that helps. I was wondering if there's a simple way I could modify the input file from the psml pseudos to generate ones that do work with SOC?

Thanks,

Danny Bennett

ebousquet
Posts: 469
Joined: Tue Apr 19, 2011 11:13 am
Location: University of Liege, Belgium

Re: Reading psml pseudodojo potentials with spin-orbit

Post by ebousquet » Wed Nov 18, 2020 9:00 am

Dear Danny,
It is confirmed that the psml+soc is not interfaced in Abinit, I've put this issue in Git with a clear reply:
https://gitlab.abinit.org/trunk/abinit/-/issues/103

Thank you for reporting the problem.

In the mean time, do you really need to read a psml, are they from pseudodojo? If yes you can use the psp8 format which eaxtly the same as the psml ones but if you want to use a psml other than pseudodojo, then it is not possible at the moment...

Best wishes,
Eric

dbennett1994
Posts: 7
Joined: Sat Mar 21, 2020 10:21 am

Re: Reading psml pseudodojo potentials with spin-orbit

Post by dbennett1994 » Fri Nov 20, 2020 11:35 am

Dear Eric, thanks for letting me know, I wanted to be sure that it wasn't something I was doing. I am happy enough to use the psp8 from pseudodojo for now.

Danny Bennett

Locked