Could not repeat the code in the arXiv:2305.11354 paper. about pyrovskite HOT 3 CLOSED

I found the solution in the Jupyter Notebook file (The_builder_module.ipynb in the example directory).

First of all, I need to use ase.io.read() function to load the .xyz file.

>>> import ase
>>> 
>>> Ap = ase.io.read("./methylammonium.xyz")
>>> 
>>> in_A_2drp = make_2drp(Ap, "Cs", "Pb", "I", 2, 3)
>>> in_A_2drp.write("AAAA.cif")

It works now!

I have an additional question. I understand the second to last parameter of “2” means two layers, but I am not sure what's the meaning of the last parameter "3" stand for? The comments in In [5] showed that it is "2*BX-bond length", while in In [6] it is presented as "BX-bond length", which makes me confused....

from pyrovskite.

First off thanks very much for using the code! I'm glad the A, Ap issue got sorted out, maybe I should add the ability to simply pass those filenames as strings and have the code open up the structure files automatically if it sees they're not passed as Atoms objects. That would make the first example you sent would work out as is, and it may be nice to have both methods available.

As for the make_X functions, yes the last parameter in those is the BX-bond distance. It used to be BX*2 in previous versions of the code, but before publishing the package I just changed everything to BX-bond distances, removing the factor of 2 to make it more intuitive. I thought I caught all of the spots I mentioned it in the examples but I missed the one you pointed out, so I'll update that shortly.

Also shortly I'll be adding a bunch of organic A' and A site cations to a res directory of some sort in the repo so that should make it less of a hassle having to track those down in the folder for the test data. Thanks again!

from pyrovskite.

First off thanks very much for using the code! I'm glad the A, Ap issue got sorted out, maybe I should add the ability to simply pass those filenames as strings and have the code open up the structure files automatically if it sees they're not passed as Atoms objects. That would make the first example you sent would work out as is, and it may be nice to have both methods available.

As for the make_X functions, yes the last parameter in those is the BX-bond distance. It used to be BX*2 in previous versions of the code, but before publishing the package I just changed everything to BX-bond distances, removing the factor of 2 to make it more intuitive. I thought I caught all of the spots I mentioned it in the examples but I missed the one you pointed out, so I'll update that shortly.

Also shortly I'll be adding a bunch of organic A' and A site cations to a res directory of some sort in the repo so that should make it less of a hassle having to track those down in the folder for the test data. Thanks again!

Thank you for your quick response!
I think you do not need to modify the A, Ap issue in the current code (to avoid introducing bugs), just modifying the paper is enough.

Hoping for new features. :)

from pyrovskite.