Incorrect orbital–atom assignment in TB2J for NiI2

[Cengizhan Gülhan]

Hello, I'm trying to learn how to extract magnetic exchange parameters using TB2J. While I have successfully processed CrI3 using the example, I am running into an issue with NiI2. Even though my Wannier function spreads are low (in the 0-2 Angstrom range), TB2J keeps assigning the Ni d-orbitals to the Iodine atoms, resulting in the error: "Atom 0 is magnetic but no orbitals have been assigned." I would appreciate any help on how I can prevent this. I have attached my output files and POSCAR for reference. I have checked out the other similar problems in the google groups but they didn't quite helped)

 assigned_basis.txt

 INCAR

 wannier90.win

 wannier90.wout

 wannier90_centres.xyz

 wannier90_hr.dat

 wannier90_tb.dat

 wannier90_wsvec.dat

[Cengizhan Gülhan]

Also I used the following TB2J run script (i also tried --groupby orbital/spin): 
wann2J.py --posfile POSCAR --elements Ni --index_magnetic_atoms 1  --spinor   --efermi -1.684 --kmesh 16 16 1 

Cengizhan Gülhan <cgulh...@gmail.com>, 30 Ara 2025 Sal, 03:45 tarihinde şunu yazdı: