Atomic fraction as float

[James-Drewitt]

@bjheinen Currently atomic fraction n must be an integer. Can this be made a float? For example, I have GaSn alloy with 8.5 at. % Sn. To get around this, if I multiply atomic fractions by 100 I get the following error:

Traceback (most recent call last):
File "C:\Users\James\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\LiquidDiffract\gui\optim_ui.py", line 197, in on_click_calc_sq
self.data['sq_y'] = core.calc_structure_factor(self.data['cor_x_cut'],
File "C:\Users\James\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\LiquidDiffract\core\core.py", line 562, in calc_structure_factor
avg_scattering_f = calc_average_scattering(composition, Q_cor)
File "C:\Users\James\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\LiquidDiffract\core\core.py", line 224, in calc_average_scattering
np.sum(np.asarray([x*y for x in atomic_ff for y in atomic_ff]), 0)
numpy.core._exceptions._ArrayMemoryError: Unable to allocate 11.7 GiB for an array with shape (1000000, 1570) and data type float64

[bjheinen]

This has now been added as of v1.2.4 and will be included in the next release version. For Faber-Ziman S(Q) this is fine as the normalisation to total number of atoms makes e.g. Ga_0.915,Sn_0.085 identical to Ga₉₁₅,Sn₈₅. I have simply allowed float input in the composition input table. Ashcroft-Langreth structure factors are normalised in a different way, and really only work for molecular materials, so if using this mode the input is restricted to integers.