ABINIT Discussion Forums

Hello,

I'm running abinit v 7.4.3 and trying to use GPAW setups to do some GGA calculations. However, I'm running into some problems getting abinit to read in the GPAW XML data correctly. Abinit identifies them as XML PAW data, and will do numerical calculations, but with lots of NaN results. I also get the following warning

psp7calc: WARNING -
Max. radius for Vloc was too large (> 20.00 a.u.) !
Numeric noise was possible.
Mesh size for Vloc has been set to 589.
Local potential is in "Vbare" format...
Compensation charge density is not taken into account in XC energy/potential
pspatm: atomic psp has been read and splines computed

I gunzipped the GPAW setups to have ASCII files for each of the PAW datasets. Do I perhaps need to point abinit to the GPAW *.basis files as well?
Anyone have any experience using the GPAW setups in abinit?

I'm getting the same, or at least a very similar, error without using GPAW.

I'm using JTH-PBE-atomicdata-0.2, paw data files. What is the warning coming from? And how serious is this?


Here it is:

pawpsp_calc: WARNING -
Max. radius for Vloc was too large (> 20.00 a.u.) !
Numeric noise was possible.
Mesh size for Vloc has been set to 1332.
Compensation charge density is not taken into account in XC energy/potential
pspatm: atomic psp has been read and splines computed

I have encountered the exact same problem with the GGA atomic data for K and Br.

Is this data set supposed to be carefully tested? Or what did I miss in the PAW calculation? I am following the tutorial and did everything as was told.

What version of abinit do you use ?
Do you use JTH table ?
Can you post the configuration of abinit :
In you log file from
=== Build Information ===
until the end of the HAVE_* echo
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Jordan wrote:What version of abinit do you use ?
Do you use JTH table ?
Can you post the configuration of abinit :
In you log file from
=== Build Information ===
until the end of the HAVE_* echo
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

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

CC_GNU CXX_GNU FC_GNU

HAVE_DFT_ATOMPAW HAVE_DFT_BIGDFT HAVE_DFT_LIBXC

HAVE_DFT_WANNIER90 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_INT_QUAD HAVE_FC_IOMSG HAVE_FC_ISO_C_BINDING

HAVE_FC_LONG_LINES HAVE_FC_MOVE_ALLOC HAVE_FC_PRIVATE

HAVE_FC_PROTECTED HAVE_FC_STREAM_IO HAVE_FC_SYSTEM

HAVE_FFT HAVE_FFT_FFTW3 HAVE_FFT_MPI

HAVE_FFT_SERIAL HAVE_FORTRAN2003 HAVE_GW_DPC

HAVE_LIBPAW_ABINIT HAVE_LINALG HAVE_LINALG_SERIAL

HAVE_MPI HAVE_MPI2 HAVE_MPI_IO

HAVE_MPI_TYPE_CREATE_S... HAVE_OS_LINUX HAVE_TIMER

HAVE_TIMER_ABINIT HAVE_TIMER_MPI HAVE_TIMER_SERIAL

USE_MACROAVE
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Here you are. Anything suspicious? I did check the XML file and did not see a suspicious radius. Instead, it is something like <paw_radius rc=" 2.2000000000"/>.