yambo 4.2.0 - NetCDF: Variable not found

[Daniele Varsano]

Dear Kushal,
It is not clear to me your answer to Davide.
In your previous post you wrote:

I moved these files to the original SAVE and it is now reading these files but I get the following error.

and this is not allowed. You need to run yambo in the folder FixSave/SAVE

In any case if can always start from scratch by doing a qe calculation and use the p2y interface. In this case you will not have format problem.

Daniele

[gyro555in]

I tried both ways to be honest and the result ended up being the same. It is accessing the new wf files.

[Davide Sangalli]

Yeah, sorry for the silly question.
I just tried the hBN tutorial with the 4.2 version and it works just fine.
From the error message you get it seems there maybe a problem with the blas libraries.
Can you try to also attach the file report in the folder config generated after the compilation of yambo ?

Best,
D.

[gyro555in]

Well the compilation is done in our central cluster so I am gonna ask my admin and get back.

[gyro555in]

Here is the compilation report used on our cluster.

#
# [VER] 4.2.0 r.109
#
# - GENERAL CONFIGURATIONS -
#
# [SYS] linux@x86_64
# [SRC] /opt/install/simulation/yambo-4.2.0/yambo-git2/yambo
# [TGT] /opt/install/simulation/yambo-4.2.0/yambo-git2/yambo
# [BIN] /opt/install/simulation/yambo-4.2.0/yambo-git2/yambo/bin
# [LIB] /opt/install/simulation/yambo-4.2.0/yambo-git2/yambo/lib/external
#
# [EDITOR] vim
# [ MAKE ] make
#
# [-] Double precision
# [X] Keep object files
# [X] Run-Time timing profile
# [-] Run-Time memory profile
#
#
# - PARALLEL SUPPORT -
#
# [X] MPI
# [-] OpenMP
#
# - LIBRARIES [E=external library; I?=internal library (c=to be compiled / f=found already compiled); X=system default; -=not used;] -
#
# I/O LIBS
#
# [ E ] IOTK : /opt/install/analysis/espresso/espresso-4.3.2/S3DE/iotk/src/libiotk.a (QE qexml)
# -I/opt/install/analysis/espresso/espresso-4.3.2/S3DE/iotk/src
# [ - ] ETSF_IO :
#
# [ Ic] NETCDF : -L/opt/install/simulation/yambo-4.2.0/yambo-git2/yambo/lib/external/intel/mpifort/v3/lib -lnetcdff -L/opt/install/simulation/yambo-4.2.0/yambo-git2/yambo/lib/external/intel/mpifort/v3/lib -lnetcdf -lm (With large files support)
# -I/opt/install/simulation/yambo-4.2.0/yambo-git2/yambo/lib/external/intel/mpifort/v3/include -I/opt/install/simulation/yambo-4.2.0/yambo-git2/yambo/lib/external/intel/mpifort/v3/include
# [ - ] HDF5 :
#
#
# MATH LIBS
#
# [ Ic] FFT : -L/opt/install/simulation/yambo-4.2.0/yambo-git2/yambo/lib/external/intel/mpifort/lib -lfftw3 (Internal FFTW3)
# -I/opt/install/simulation/yambo-4.2.0/yambo-git2/yambo/lib/external/intel/mpifort/include/
# [ E ] BLAS : -lblas
# [ E ] LAPACK : -llapack
# [ - ] SCALAPACK :
# [ - ] BLACS :
# [ - ] PETSC :
#
# [ - ] SLEPC :
#
#
# OTHER LIBS
#
# [ Ic] LibXC : -L/opt/install/simulation/yambo-4.2.0/yambo-git2/yambo/lib/external/intel/mpifort/lib -lxcf90 -lxc
# -I/opt/install/simulation/yambo-4.2.0/yambo-git2/yambo/lib/external/intel/mpifort/include
# [ E ] MPI : -L/opt/pkg/mpi/openmpi/1.10.2/intel/15.2/64/opt/lib -lmpi
# -I/opt/pkg/mpi/openmpi/1.10.2/intel/15.2/64/opt/include
#
# - COMPILERS (FC kind= intel) -
#
# [ CPP ] /opt/pkg/compiler/intel/fc/15.2/bin/icpc -E -g0 -O3 -m64 -gcc-name=gcc -gxx-name=g++ -cxxlib=/opt/pkg/compiler/gnu/gcc/4.8.2/lib64 -D_MPI -D_FFTW -D_TIMING
# [ FPP ] fpp -free -P -D_MPI -D_FFTW -D_TIMING
# [ CC ] mpicc -g -O2 -D_C_US -D_FORTRAN_US
# [ FC ] mpifort -assume bscc -g -O3 -ip
# [ FCUF] -assume bscc -g -O0
# [ F77 ] mpifort -assume bscc -g -O3 -ip
# [ F77U] -assume bscc -g -O0
# [Cmain] -nofor_main

[Davide Sangalli]

Ok. So blas and lapack are the only ones external.
Ask if they can re-install the code with the option
./configure --enable-int-linalg

Also I would suggest the use of the option:
--with-extlibs-path=$SOME_PATH

Yambo will install all the needed libraries there once for all and eventually re-use them in the future.

Best,
D.

[gyro555in]

Your suggestions for the configuration worked and I tested some basic input files. Thank you very much for your support.

[Alesster]

I am writing here because my problem is connected with this topic.

I use QE v6.3 and Yambo v 4.2.1
Yambo was compiled with --with-p2y-version=qexsd

I am doing simple example Silicon. When I am calculating optical properties it gives error:
[ERROR] STOP signal received while in :[04] Optics
[ERROR] File ./SAVE//ndb.dip_iR_and_P; Variable NOT DEFINED; NetCDF: HDF error

What is the problem can be?

Thank you,
Best,
Sasha

[Daniele Varsano]

Dear Aleksandr,

I use QE v6.3 and Yambo v 4.2.1
Yambo was compiled with --with-p2y-version=qexsd

Ok, if p2y runned smoothly, all is ok up to here.

When I am calculating optical properties it gives error:
[ERROR] STOP signal received while in :[04] Optics
[ERROR] File ./SAVE//ndb.dip_iR_and_P; Variable NOT DEFINED; NetCDF: HDF error
What is the problem can be?

Here we need to investigate, and reproduce your problem. Can you also post your config.log and ./config/setup files.
Before that please note that in your input file there is a problem:

Code: Select all

% XfnQP_E
 0.530000 | 1.000000 | 1.000000 |        # [EXTQP Xd]

There is a missing % closing the filed, it should be:

Code: Select all

% XfnQP_E
 0.530000 | 1.000000 | 1.000000 |        # [EXTQP Xd]
%

May be you can try if this solve your problem. Next please note that with these settings:

Code: Select all

FxcGRLc= 1             RL      # [TDDFT] XC-kernel RL size
NGsBlkXd= 1   

you essentially not including local field effects ant the ALDA kernel is not properly evaluated, but this is not at all the reason of the error you get.

Best,
Daniele

[Alesster]

Dear Daniele,

Thank you for the reply.

The config.log is very huge therefore I separated it on two parts.

if p2y runned smoothly, all is ok up to here.

Yes, p2y finished correctly.

There is a missing % closing the filed

In original file % is not missing.

you essentially not including local field effects ant the ALDA kernel is not properly evaluated

This is just test calculation and it must not influence on the running process. I did such kind of runs before and it was always ok.

Best,
Sasha.