[Feature] Standardise the order the axes are listed in arguments

[IgorTatarnikov]

Currently, parts of the package list arguments using the xyz order which are then flipped to match the standard zyx. To avoid confusion and bugs, all parameters relating to the physical axes (not anatomical) should always be in the zyx order, and passed using a tuple where possible.

e.g. The below code snippet is confusing, and has a bug inside as the signal emits the scales in xyz order while the atlas resolutions are in zyx. The test only passes due to the symmetric nature of the atlas.

brainglobe-registration/tests/test_registration_widget.py

Lines 152 to 175 in 027493c

	def test_scale_moving_image_2d(
	make_napari_viewer_with_images,
	registration_widget,
	mocker,
	x_scale_factor,
	y_scale_factor,
	):
	mock_atlas = mocker.patch(
	"brainglobe_registration.registration_widget.BrainGlobeAtlas"
	)
	mock_atlas.resolution = [20, 20, 20]
	registration_widget._atlas = mock_atlas
	current_size = registration_widget._moving_image.data.shape
	registration_widget.adjust_moving_image_widget.scale_image_signal.emit(
	mock_atlas.resolution[1] * x_scale_factor,
	mock_atlas.resolution[2] * y_scale_factor,
	0.001,
	)
	assert registration_widget._moving_image.data.shape == (
	current_size[0] * y_scale_factor,
	current_size[1] * x_scale_factor,
	)

[alessandrofelder]

Would it be helpful to move away from zyx/xyz altogether (I think there was a plan to somewhere, but best I can find is: brainglobe/cellfinder#413) and use axis 0/1/2 or terms like "in-plane" and "axial"? not sure 🤔

[adamltyson]

I was going to add this, here's the issue: brainglobe/brainglobe-utils#37