Data linking in ApplicationState._link_new_data is too specific to live in Application

[astrofrog]

The following code in Application:

jdaviz/jdaviz/app.py

Lines 241 to 260 in d296c63

	def _link_new_data(self):
	"""
	When additional data is loaded, check to see if the spectral axis of
	any components are compatible with already loaded data. If so, link
	them so that they can be displayed on the same profile1D plot.
	"""
	new_len = len(self.data_collection)
	# Can't link if there's no world_component_ids
	wc_new = self.data_collection[new_len-1].world_component_ids
	if wc_new == []:
	return
	# Link to the first dataset with compatible coordinates
	for i in range(0, new_len-1):
	wc_old = self.data_collection[i].world_component_ids
	if wc_old == []:
	continue
	else:
	self.data_collection.add_link(LinkSame(wc_old[0], wc_new[0]))
	break

is actually wrong for some applications such as MOSViz. It will link the first world coordinates of all datasets, regardless of whether the data are images or spectra and so on. This means that some links will end up being Declination and Spectral axis and so on.

I think this code made sense for e.g. specviz and probably cubeviz but it is too specific to live in the main Application (and is incorrect for some applications)

[pllim]

Since we actually store the config, probably not hard to only run this if certain viz is loaded:

jdaviz/jdaviz/app.py

Lines 1205 to 1206 in d296c63

	# store the loaded config object
	self._loaded_configuration = config

[astrofrog-conda-forge]

I don't think we want to have it live on the main class and optionally select it based on the configuration - after all that is what the subclasses are for - so it could live on the relevant applications. I don't think the default application should do any auto linking.

[pllim]

I see people calling Application(configuration) in their notebooks. I don't think subclass will work.

[astrofrog]

Ah I see - then yes maybe there should be an option for this, but in any case the way the function is written is prone to breakage so needs to be improved to more explicitly link spectral coordinates. I am writing up an issue about improving linking across the board, I will touch on that function in there.

[pllim]

xref #804