Parity in final state

[redeboer]

In ampform-dpd, the final and initial state are defined with Particles. They require spin, parity, mass and width.

In the latest model example, the initial state and final state contain only spin and mass – understandable, as they are under the heading kinematics.

amplitude-serialization/Lc2ppiK.json

Lines 8 to 33 in 9fcd26e

	"initial_state": {
	"index": 0,
	"name": "Lc",
	"spin": "1/2",
	"mass": 2.28646
	},
	"final_state": [
	{
	"index": 1,
	"name": "p",
	"spin": "1/2",
	"mass": 0.938272046
	},
	{
	"index": 2,
	"name": "pi",
	"spin": "0",
	"mass": 0.13957018
	},
	{
	"index": 3,
	"name": "K",
	"spin": "0",
	"mass": 0.493677
	}
	]

Is the inclusion of parity redundant? I was thinking of computing the parities from the parity_factor in the vertices, but am not sure how and it is a bit of a pain to search through the vertices to get info about the kinematics.
Could it be that the inclusion of parity and width in ampform-dpd was a design mistake? Width in final state particles is indeed redundant in most cases, unless you want to parametrize mass.

[redeboer]

Outcome from the discussion: parity is not needed for the formulation of the model if parity_prefactor and (optionally) LS values are provided. Parity is only needed for amplitude builders (i.e. compute allowed LS couplings), but the purpose of the amplitude serialization format is to be a minimal and complete set of info to formulate a mathematical model, independent of building procedure.