Some confusion on the description in the joss paper of this project.

[hongyi-zhao]

Some confusion on the description in the joss paper of this project.

[hongyi-zhao]

If I understand correctly, your main point is as follows:

Your representation is not obtained from a table lookup but is directly calculated using a specific algorithm. Moreover, this algorithm also provides the corresponding basis, and that's all there is to it. In fact, for people doing physics, it doesn't matter how the representation is derived; what matters more is having the appropriate representation and knowing how to use it.

Am I right?

Regards,
Zhao

[lan496]

Thank you for your comment on the tiny description.

1. I agree this sentence is confusing. I mean irreps can be calculated by Bloch wave functions for each eigenenergy, and the construction of irreps itself does not rely on other tabulations.
2. Of course we can choose any vector spaces with the irreps's dimensions. Spgrep simply provides representation matrices calculated in a deterministic algorithm.
3. I mean symmetry-adapted basis constructed by the unique irreps. In total, we say the uniqueness of irreps up to representation matrices.

[lan496]

Yes, this package only concerns representation matrices.