Order matters for checking equality of structure about pymatgen HOT 5 CLOSED

@bpuchala Would you be able to submit a fix for this that implements the type promotion so we can see if there are any unintended downstream implications?

Sorry, I'm just a user here :)

from pymatgen.

Oh dear, loss of symmetry in __eq__ is pretty serious, esp. in the core Structure class.
@bpuchala Thanks for reporting and thanks for finding the root cause! 🙏

I'm in favor of making Structure.__eq__ more discriminative by promoting Elements to Species with oxi state 0 so that

element_structure == species_structure
>>> False
# and
species_structure == element_structure
>>> False

but I'm interested to hear @shyuep and @mkhorton's opinions.

@bpuchala Would you be able to submit a fix for this that implements the type promotion so we can see if there are any unintended downstream implications?

from pymatgen.

I would just fix Element such that it only equates equal to Element and not Specie. I dislike the ElementBase implementation. I think it overcomplicates matters relative to just a straight Enum.

from pymatgen.

I would just fix Element such that it only equates equal to Element and not Specie.

That's already the case.

from pymatgen.core import Element, Species

Element('O') == Species('O')
>>> False

I dislike the ElementBase implementation. I think it overcomplicates matters relative to just a straight Enum.

I've said the same thing in the past. IIRC @mkhorton had some reservations about that.

from pymatgen.

@janosh No, that wasn't the case. Species("Fe", 0) == Element("Fe") but Element("Fe") != Species("Fe", 0). That was the whole crux of the issue reported. I already pushed a fix.

from pymatgen.