Implement ElectricMultipoleParameters

[EZoni]

Close #38.

To do:

• Implement new class ElectricMultipoleParameters.
• Add new tests and make existing tests more symmetric.

[EZoni]

@ax3l @cemitch99

I was looking into duplicating most of the MagneticMultipoleParameters tests for the ElectricMultipoleParameters class, and I ran into a question related to the Quadrupole class.

Is it expected that the ElectricMultipoleP attribute is optional, while the MagneticMultipoleP attribute is mandatory?

pals-python/src/pals/kinds/Quadrupole.py

Lines 7 to 15 in 2ddf2a7

	class Quadrupole(ThickElement):
	"""Quadrupole element"""
	# Discriminator field
	kind: Literal["Quadrupole"] = "Quadrupole"
	# Octupole-specific parameters
	ElectricMultipoleP: Optional[ElectricMultipoleParameters] = None
	MagneticMultipoleP: MagneticMultipoleParameters

Shouldn't a quadrupole be either electric or magnetic?

[cemitch99]

@EZoni Here is my point of view: By far the most common usage of "Quadrupole" refers to a magnetic quadrupole, for which all the electric multipole contributions vanish. However, there has been a general trend to allow optional contributions from other parameter groups where reasonable (for example, to support different types of "combined-function" elements). There are also pure electric quadrupoles, for which the same situation applies in reverse (although these are less common). If we wanted to treat electric and magnetic quadrupoles in a symmetric way, there would be two options: 1) create a separate ElectricQuadrupole kind; or 2) require the specification of either magnetic or electric quadrupole components or both (e.g., at least one must be specified). Either option appears reasonable to me.

[EZoni]

Thanks, @cemitch99!

The standard, https://pals-project.readthedocs.io/en/latest/element-kinds.html#quadrupole-element, seems to go more in the direction of your option 2, "require the specification of either magnetic or electric quadrupole components or both (e.g., at least one must be specified)", since it doesn't mention any ElectricQuadrupole kind.

I think I will propose a change that implements option 2 and see if it works for everyone.

[cemitch99]

@EZoni That sounds good; I agree it's a simple and clean option.

[EZoni]

@cemitch99

Do you think the same is true for Sextupole, Octupole, and Multipole?

Right now they seem to allow for no parameter at all, neither MagneticMultipoleP nor ElectricMultipoleP, i.e., both are optional.

For example, we have this

• pals-python/src/pals/kinds/Sextupole.py

  Lines 15 to 17 in 2ddf2a7

  	# Sextupole-specific parameters
  	ElectricMultipoleP: Optional[ElectricMultipoleParameters] = None
  	MagneticMultipoleP: Optional[MagneticMultipoleParameters] = None

• pals-python/src/pals/kinds/Octupole.py

  Lines 15 to 17 in 2ddf2a7

  	# Octupole-specific parameters
  	ElectricMultipoleP: Optional[ElectricMultipoleParameters] = None
  	MagneticMultipoleP: Optional[MagneticMultipoleParameters] = None

• pals-python/src/pals/kinds/Multipole.py

  Lines 15 to 17 in 2ddf2a7

  	# Multipole-specific parameters
  	ElectricMultipoleP: Optional[ElectricMultipoleParameters] = None
  	MagneticMultipoleP: Optional[MagneticMultipoleParameters] = None

without extra validation logic that requires at least one of the two to be present (like the logic I added in this PR for Quadrupole).

I could fix this in a separate PR, once the "under construction" flag is removed. Do you think this does need a fix indeed?

[cemitch99]

@EZoni I plan to contribute Sextupole, Octupole, and Multipole sections to the standard shortly, since they are simple and widely used. I don't see any good argument for allowing these element kinds with no multipole components at all (since then they have no reason to be classified this way), so adding a validation doesn't hurt. In some ways, it is a bare minimum sanity check.

[ax3l]

Oh interesting, is this a PALS defect that we have no normalized parameters for the ElectricMultipoleParameters?
https://pals-project.readthedocs.io/en/latest/element-parameters.html#electricmultipolep-electric-multipole-parameters

pals-python/src/pals/parameters/MagneticMultipoleParameters.py

Lines 33 to 34 in 35ed73a

	- KnN: Normalized normal component of Nth order multipole
	- KsN: Normalized skew component of Nth order multipole

I would expect this to be identical as for the magnetic parameters.

[DavidSagan]

I've never seen normalized electric field strengths used which is why I left them out. If someone actually does use them then they should be put in.

[ax3l]

Ok, perfect! Thanks for checking David! 🙏

[ax3l]

LGTM. As a follow-up, please check if normalized is missing in the PALS standard for electric multipole params. Looks like an oversight/inconsistency to me.