kernels

[Jacfomg]

Prototype routine to test the kernel computation for integration using zhinst functions.
Also, results.npz and result.json will be generated, which I didn't do correctly and all routines crash.

Tests will fail until we merge qiboteam/qibolab#752 to get kernels in the dummy.

Checklist:

• Reviewers confirm new code works as expected.
• Tests are passing.
• Coverage does not decrease.
• Documentation is updated.
• Compatibility with Qibo modules (Please edit this section if the current pull request is not compatible with the following branches).
  • Qibo: master
  • Qibolab: main
  • Qibolab_platforms_qrc: main

[andrea-pasquale]

Thanks @Jacfomg for implementing this.
I still need to run the protocol.
The modifications in the API look good to me.

[andrea-pasquale]

Can we move this directly in the __init__?
With default value None

[Jacfomg]

No, because then it was complaning that I was giving after other non default arguments. I mean, we should do something else if we want it to get it out of there.

[Edoardo-Pedicillo]

If possible, try to implement the same approach of the Parameters class, to be consistent with the rest of the code.

qibocal/src/qibocal/auto/operation.py

Lines 76 to 98 in 6d3de62

	@classmethod
	def load(cls, input_parameters):
	"""Load parameters from runcard.
	Possibly looking into previous steps outputs.
	Parameters defined in Parameters class are removed from `parameters`
	before `cls` is created.
	Then `nshots` and `relaxation_time` are assigned to cls.
	.. todo::
	move the implementation to History, since it is required to resolve
	the linked outputs
	"""
	default_parent_parameters = deepcopy(DEFAULT_PARENT_PARAMETERS)
	parameters = deepcopy(input_parameters)
	for parameter, value in default_parent_parameters.items():
	default_parent_parameters[parameter] = parameters.pop(parameter, value)
	instantiated_class = cls(**parameters)
	for parameter, value in default_parent_parameters.items():
	setattr(instantiated_class, parameter, value)
	return instantiated_class

[Jacfomg]

Is this what @GabrielePalazzo did some time ago ? It was not very clear to me what it was doing could you lend me a hand ?

[Edoardo-Pedicillo]

You can also try something easier like

qibocal/src/qibocal/auto/operation.py

Lines 79 to 82 in eca6567

	nboot = parameters.pop("nboot", 0)
	par = cls(**parameters)
	par.nboot = nboot
	return par

[Edoardo-Pedicillo]

Is this PR in draft or under review?

[Jacfomg]

Well, it was draft because the results were not the expected, I'm waiting for comments from Zurich. But the changes to the API can be reviewed and maybe put them on another PR so they get merged faster.

[Edoardo-Pedicillo]

If possible, try to implement the same approach of the Parameters class, to be consistent with the rest of the code.

qibocal/src/qibocal/auto/operation.py

Lines 76 to 98 in 6d3de62

	@classmethod
	def load(cls, input_parameters):
	"""Load parameters from runcard.
	Possibly looking into previous steps outputs.
	Parameters defined in Parameters class are removed from `parameters`
	before `cls` is created.
	Then `nshots` and `relaxation_time` are assigned to cls.
	.. todo::
	move the implementation to History, since it is required to resolve
	the linked outputs
	"""
	default_parent_parameters = deepcopy(DEFAULT_PARENT_PARAMETERS)
	parameters = deepcopy(input_parameters)
	for parameter, value in default_parent_parameters.items():
	default_parent_parameters[parameter] = parameters.pop(parameter, value)
	instantiated_class = cls(**parameters)
	for parameter, value in default_parent_parameters.items():
	setattr(instantiated_class, parameter, value)
	return instantiated_class

[Jacfomg]

The calculation of the kernel calculation was moved https://docs.zhinst.com/labone_q_user_manual/release_notes/

[Jacfomg]

It needs laboneq 2.20.0 version that will get merged soon. And testing on hardware.

[andrea-pasquale]

Thanks @GabrielePalazzo for the review. Your comments have been addressed.