settings in input.kpt not used

[wangwy13]

Dear developers,

I have compiled v1.8.2 with the following options:
../Installer.py build -c CC=mpicc CXX=mpicxx FC=mpif90 FCFLAGS='-O2 -fopenmp -L /opt/cuda/lib64' CFLAGS='-fopenmp' CXXFLAGS='-fopenmp' --enable-cuda-gpu --with-cuda-path=/opt/cuda --with-ext-linalg='${MKLROOT}/lib/intel64/libmkl_scalapack_lp64.a -Wl,--start-group ${MKLROOT}/lib/intel64/libmkl_intel_lp64.a ${MKLROOT}/lib/intel64/libmkl_intel_thread.a ${MKLROOT}/lib/intel64/libmkl_core.a ${MKLROOT}/lib/intel64/libmkl_blacs_intelmpi_lp64.a -Wl,--end-group -liomp5 -lpthread -lm -ldl -lstdc++' NVFLAGS="-O3 -arch sm_35"

The building is successful.

Strangely, the settings in input.kpt appears to be not used.
For example, I have a input.kpt in the directory with,

MPgrid
3 3 3
1
0.0 0. 0.0

however, I got the following

 kpt:
   method : manual # K-point sampling method
   kpt: # Kpt coordinates
   - [0., 0., 0.]
   wkpt: [1.] # Kpt weights

[Luigi Genovese (luigi-genovese)]

Hello,

many thanks for the message.

The problem you have is related to the modifications of the input scheme of the code
of the 1.8.x series.
Unfortunately, some of the webpages of the BigDFT site are outdated since more than three years. We are working on finding a more manageeable version of distributing the modification in the usage of the code.

The conversion to the new input format is automatically done with the bigdft-tool: run the bigdft-tool script with the old set of input files inside and you will see the input.yaml file generated.

Do not hesitate

Many thanks

Luigi

[wangwy13 (wangwy13)]

Dear Luigi,
Many thanks for your suggestion and your efforts. Everything is fine now!

Yi Wang