Will HamGNN support fitting the TB Hamiltonian generated by ABACUS in the future?

[sgyang345]

Dear Developer,

I would like to express my admiration for the outstanding work you've done!

To the best of my knowledge, there is a first-principles software known as ABACUS that boasts support for numerical atomic orbital basis sets. ABACUS is equipped with a wealth of features, precision in calculations, and a rapid pace of development. It's no surprise that an increasing number of people are turning to it. I'm eagerly anticipating the opportunity to use it as the basis for HamGNN in fitting the TB Hamiltonian. We kindly request the developers' support in enabling seamless integration with ABACUS.

Thank you for your dedication and support!

Warm regards, SgYang

[QuantumLab-ZY]

Dear SgYang,

Thank you very much for your valuable feedback. Currently, HamGNN can also be used to fit the tight-binding Hamiltonian generated by ABACUS. However, the scripts for processing ABACUS Hamiltonian are not yet ready for upload. We will prepare these scripts in the next few days and inform you here when we upload them.

Best wishes, Yang Zhong

[QuantumLab-ZY]

Dear SgYang,

Sorry, because I've been quite busy for the past several months, I haven't had the chance to test and release the supporting tools for ABACUS. I've just uploaded the script for handling the ABACUS Hamiltonian to the utils_abacus directory. Next, I'll provide more detailed instructions on its usage in the readme. If you're interested in fitting the Hamiltonian matrix of ABACUS using HamGNN, you can give these tools a try.

Best wishes, Yang Zhong

[sgyang345]

Thank you for your work and reply, I think this will benefit many people in need.

[Youhaojen]

Dear Yang Zhong,

Thanks for your nice work. I tried the MD example with the modified ABACUS. Here, the following error happened.

 !! CONVERGENCE HAS NOT BEEN ACHIEVED !!

===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   RANK 0 PID 249 RUNNING AT 2a17f5f8342f
=   KILLED BY SIGNAL: 9 (Killed)
===================================================================================

===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   RANK 1 PID 250 RUNNING AT 2a17f5f8342f
=   KILLED BY SIGNAL: 11 (Segmentation fault)
===================================================================================

Is the modified ABACUS not supporting the MD method?

P.s The Dockerfile can be used as reference. Dockerfile.zip

Thanks

Hao-Jen You

[QuantumLab-ZY]

Dear Hao-Jen You,

The modified version of ABACUS is only used for exporting Hamiltonians H0 that do not need to be calculated by SCF iterations. The modified version of ABACUS cannot be used for any other calculations. You need to use a regular ABACUS to perform SCF calculations or other calculations.

Best wishes, Yang Zhong

Dear Yang Zhong,

Thanks for your nice work. I tried the MD example with the modified ABACUS. Here, the following error happened.

 !! CONVERGENCE HAS NOT BEEN ACHIEVED !!

===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   RANK 0 PID 249 RUNNING AT 2a17f5f8342f
=   KILLED BY SIGNAL: 9 (Killed)
===================================================================================

===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   RANK 1 PID 250 RUNNING AT 2a17f5f8342f
=   KILLED BY SIGNAL: 11 (Segmentation fault)
===================================================================================

Is the modified ABACUS not supporting the MD method?

P.s The Dockerfile can be used as reference. Dockerfile.zip

Thanks

Hao-Jen You