Update GST to take in gates.Unitary

[mho291]

Checklist:

• Reviewers confirm new code works as expected.
• Tests are passing.
• Coverage does not decrease.
• Documentation is updated.

[renatomello]

Suggested change

	if isinstance(params, list):
	init_args = signature(gate).parameters
	valid_angles = [arg for arg in init_args if arg in angles]
	angle_values = dict(zip(valid_angles, params))
	else:
	init_args = signature(gate).parameters
	valid_angles = [arg for arg in init_args if arg in matrix]
	angle_values = dict(zip(valid_angles, [params]))
	if not isinstance(params, list):
	params = [params]
	init_args = signature(gate).parameters
	valid_angles = [arg for arg in init_args if arg in angles]
	angle_values = dict(zip(valid_angles, params))

[mho291]

Thanks @renatomello for spotting this. I have made the changes. But still trying to get it to work. It's currently complaining

FAILED tests/test_tomography_gate_set_tomography.py::test_GST[numpy-False-target_gates0] - ValueError: axes don't match array
FAILED tests/test_tomography_gate_set_tomography.py::test_GST[numpy-True-target_gates0] - ValueError: axes don't match array
FAILED tests/test_tomography_gate_set_tomography.py::test_GST[qibojit-numba-False-target_gates0] - ValueError: axes don't match array
FAILED tests/test_tomography_gate_set_tomography.py::test_GST[qibojit-numba-True-target_gates0] - ValueError: axes don't match array

Will keep trying to debug.

[renatomello]

This docstring is generating a warning when the documentation is compiled because of identation. I fixed it in #1584. Please look at it to also fix it here.

[mho291]

Hi @MatteoRobbiati, I need some advice with this. When I run the pytests locally, they pass. But here, they always run into dimension error:

FAILED tests/test_tomography_gate_set_tomography.py::test_GST[qiboml-pytorch-False-target_gates0] - RuntimeError: permute(sparse_coo): number of dimensions in the tensor input does not match the length of the desired ordering of dimensions i.e. input.dim() = 1 is not equal to len(dims) = 2

Could there be an issue with qiboml-pytorch?

Thanks so much!

[MatteoRobbiati]

Could there be an issue with qiboml-pytorch?

Ehi @mho291! Thanks for asking :)
I tested locally your branch and in my case I get the same failure, but on all the backends (numpy, qibojit, pytorch, tensorflow), so ig the source of the problem is elsewhere.
It seems to me the problem is appearing when calling this line in gates.py.