Possible bugs about GTO gradient calculation and Bragg-Slater radius

[fate1997]

Hi Ivo,

I recently tried to follow up on your codes to learn DFT, they are very helpful for beginners, thank you!

I found there are some codes I am confused and they might be bugs:

1. Here, you use $(x-x_0)^{l-1}$ represent the gradient for the angular term, shouldn't it be $l\times (x-x_0)^{l-1}$?
2. Here in your pydft package, it seems that you get Bragg-Slater radius by directly passing the atom index, but since your Bragg-Slater radius list is started from hydrogen instead of atom placeholder, so the true way should be bragg_slater.BSRADII[self.__aidx-1].

[ifilot]

Thank you for your kind words and bringing this error to my attention.

• The l term was indeed missing in the calculation of the gradient. Furthermore, this error went undetected due to the unit test not testing for sufficiently high values for l,m,n. I have fixed the error and expanded the scope of the unit test.
• Your second point, though of course very valid, relates to a different repository. I have raised a new issue in that repository where it will also be addressed in brief time.

Thank you once more for filing this issue. It is very much appreciated.

[fate1997]

Thank you for the quick fix!