tel_id in photoelectrons object is index rather than id?

[orelgueta]

Looking at a few events, I think that the telescope id in the Photolectrons object defined below is just the index rather than the telescope id itself. That is, it starts at zero and goes to n_tel - 1.

pyeventio/src/eventio/iact/objects.py

Line 299 in 79228db

self.telescope_id = header.id % 1000

This definition is different to the definition in the TelescopeEvent object which provides the actual id:

pyeventio/src/eventio/simtel/objects.py

Lines 716 to 723 in 79228db

	self.telescope_id = self.type_to_telid(header.type)
	self.global_count = header.id
	@staticmethod
	def type_to_telid(eventio_type):
	base = eventio_type - 2200
	return 100 * (base // 1000) + base % 1000

Not a big deal, but if I am right and the fix is just replacing one definition with the other, it's probably worth it to fix (I can do it).

[maxnoe]

Looking at a few events, I think that the telescope id in the Photolectrons object defined below is just the index rather than the telescope id itself. That is, it starts at zero and goes to n_tel - 1.

This is a known issue in sim telarray, I will dig up some emails from Konrad, but the gist is, the object is correctly defined here in terms of doing the I/O of the eventio object, but simtel fills something "wrong".

I.e. the header struct defines this to be the telescope_id, which is why we parse it as such, but in case of simtel, it's actually not the same "telescope_id" as used everywhere else.

See e.g.:
https://github.com/cta-observatory/ctapipe/blob/3dc94da64dc1e96ed9316a612c8a6c67ada32af0/src/ctapipe/io/simteleventsource.py#L816-L820

and this issue / PR pair in ctapipe:

cta-observatory/ctapipe#1338
cta-observatory/ctapipe#1337

I'll leave this open here, so when we do eventio 2.0, we can include the fix here