[Bug]: Coordinate systems of variables do not get checked against known

[brosaplanella]

PyBaMM Version

23.5

Python Version

3.9.16

Describe the bug

Currently the coordinate system of a variable is not checked against the known system. This means that if the wrong argument is passed (e.g. due to a typo) it will default to "cartesian" without throwing an error.

Steps to Reproduce

The MWE would be very lengthy, but basically it can be checked that:

1. The only place where KNOWN_COORD_SYS is checked is to test meshes for the battery model (i.e. it does not actually check for wrong coordinate systems).
2. In finite_volume.py there is an else statement that means that anything not being cylindrical polar or spherical polar will default to cartesian, see

PyBaMM/pybamm/spatial_methods/finite_volume.py

Lines 139 to 150 in fffe650

	# check coordinate system
	if submesh.coord_sys in ["cylindrical polar", "spherical polar"]:
	second_dim_repeats = self._get_auxiliary_domain_repeats(symbol.domains)
	# create np.array of repeated submesh.edges
	r_edges_numpy = np.kron(np.ones(second_dim_repeats), submesh.edges)
	r_edges = pybamm.Vector(r_edges_numpy)
	if submesh.coord_sys == "spherical polar":
	out = divergence_matrix @ ((r_edges**2) * discretised_symbol)
	elif submesh.coord_sys == "cylindrical polar":
	out = divergence_matrix @ (r_edges * discretised_symbol)
	else:
	out = divergence_matrix @ discretised_symbol

Relevant log output

No response

[brosaplanella]

Ooops, that was already reported in #2220 (contains some additional info). Let's make sure we close both issues when we merge PR.

[brosaplanella]

@RuiheLi are you working on this?

[RuiheLi]

Yes, I will. Will look into #2220. Thanks for the remind! @brosaplanella

[RuiheLi]

Yes, I will. Will look into #2220. Thanks for the remind! @brosaplanella

Just checked the time of both issue, will look into both.

[rtimms]

closing, see #4600