SSCHA
Packages and their version #367
-
|
Dear all I encountered a "TypeError: float() argument must be a string or a real number, not 'Atom'" error during my calculation and the SnTe_ToyModel example. I also found the same error in problem #184. After checking, I found that my spglib version is 2.2. It seems there might be version compatibility issues with other packages, such as cellstructure and ase. Could you please list all your packages along with their versions? |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments 8 replies
-
|
I also found that it cannot read the lattice correctly when the lattice vectors are [[1, 0, 0], [0, 1, 0], [0, 0, 1]]. In this case, the three lattice vectors are incorrectly combined into a single vector [1, 1, 1]. |
Beta Was this translation helpful? Give feedback.
-
|
Hi, this seems strange. Could you please copy the full error stack, so we can reconstruct the precise portion of the code raising this issue? |
Beta Was this translation helpful? Give feedback.
-
|
Here is my packages along with their versions. |
Beta Was this translation helpful? Give feedback.
-
|
My quantum espresso version is 7.3.1 |
Beta Was this translation helpful? Give feedback.
-
|
I also found "Note: The following floating-point exceptions are signalling: IEEE_DENORMAL" during the sscha calculations. And there is no output information in frequencies.dat.dat file. I am not sure whether it would affect final results. |
Beta Was this translation helpful? Give feedback.
-
|
Hi, this is the typical spglib version error as in #184
|
Beta Was this translation helpful? Give feedback.
Hi, this is the typical spglib version error as in #184
Your python-package specify:
spglib version 2.5
while you must have 2.2 or below (they dropped support for ASE atoms).
Downgrade spglib to version 2.2 and it should work.
We will most likely fix this in a later version of cellconstructor as python-sscha by writing a wrapper, however, it is used quite diffusely in the code right now.