Regarding the unit conversion between angstroms and Bohr, the inconsistency between dpdata and ABACUS

[JGHan7]

For some calculations that require a more accurate structure, it is very convenient to convert POSCAR files to STRU files through dpdata.

And we can convert the default unit of atomic positions from Cartesian to Cartesian_angstrom, so that when LATTICE_CONSTANT is modified (for example, when a supercell is needed), it will not affect the atomic positions. However, dpdata and ABACUS are inconsistent in the unit conversion between angstroms and Bohrs, which causes a certain error in this modification.

Can this problem be improved? Thank you

By the way, one of the common results I searched on the Internet is 1A = 1.889725988579 Bohr, which is not completely consistent with dpdata and ABACUS, which is very strange...

[pxlxingliang]

Hi, @JGHan7, the setting of the same constant by different software may differ in significant digits. This should not be a problem,and the results given by any software are valid under specific accuracy.

For the problem you mentioned, you can do a test to check the effect of the inconsistent of "LATTICE_CONSTANT". Do two calculations with same LATTICE_VECTORS and ATOMIC_POSITIONS, and different "LATTICE_CONSTANT" defined in ABACUS and dpdata, and then check the final energies.

[JGHan7]

OK, I'll try it, thank you