_mat_data seems orphan

[germa89]

As the title.

Link:

pymapdl/src/ansys/mapdl/core/mapdl_grpc.py

Lines 2718 to 2746 in 1cbc037

	def _mat_data(self, pname, raw=False):
	"""Downloads matrix data from a parameter and returns a scipy sparse array"""
	try:
	from scipy import sparse
	except ImportError: # pragma: no cover
	raise ImportError("Install ``scipy`` to use this feature") from None
	minfo = self._data_info(pname)
	stype = ANSYS_VALUE_TYPE[minfo.stype]
	mtype = minfo.objtype
	shape = (minfo.size1, minfo.size2)
	if mtype == 2: # dense
	request = pb_types.ParameterRequest(name=pname)
	chunks = self._stub.GetMatData(request)
	values = parse_chunks(chunks, stype)
	return np.transpose(np.reshape(values, shape[::-1]))
	elif mtype == 3: # sparse
	indptr = self._vec_data(pname + "::ROWS")
	indices = self._vec_data(pname + "::COLS")
	vals = self._vec_data(pname + "::VALS")
	if raw: # for debug
	return vals, indices, indptr, shape
	else:
	return sparse.csr_matrix(
	(vals, indices, indptr), dtype=stype, shape=shape
	)
	raise ValueError(f'Invalid matrix type "{mtype}"')

[germa89]

It is not covered by tests either.

[germa89]

It is called by pyansys-math package...

https://github.com/search?q=repo%3Aansys%2Fpyansys-math%20_mat_data&type=code

I feel we should move it then there. Pinging @clatapie for feedback

[germa89]

Mmmhh.. naaahh.. It used quite a few MAPDL gRPC methods, then we better keep it here. But we should add some test then.

[germa89]

MMhh.. not even tests... It should be tested properly in pyansys-math. I'm closing this issue. If @clatapie wants to add tests here or in math, she is free to do it. I do not have the time!