Add ParameterAnalysis

[henrikjacobsenfys]

To fit the fit result from Analysis to either a ModelComponent, ComponentCollection or DiffusionModel.

[codecov]

Codecov Report

❌ Patch coverage is 98.87640% with 3 lines in your changes missing coverage. Please review.
⚠️ Please upload report for BASE (develop@7dbad54). Learn more about missing BASE report.

Files with missing lines	Patch %	Lines
src/easydynamics/analysis/parameter_analysis.py	98.41%	1 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             develop     #167   +/-   ##
==========================================
  Coverage           ?   98.12%           
==========================================
  Files              ?       47           
  Lines              ?     3092           
  Branches           ?      553           
==========================================
  Hits               ?     3034           
  Misses             ?       34           
  Partials           ?       24           

Flag	Coverage Δ
integration	46.89% <25.46%> (?)
unittests	98.12% <98.87%> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[rozyczko]

Well written functionality. Just a few minor comments

[rozyczko]

Why do you initialize variables as a set but then convert into a list? This potentially loses ordering (which is not really required, anywya) but is confusing. We shouldn't have any duplicates, since the unique_name mechanism prohibits that.

[henrikjacobsenfys]

Because there may be overlapping variables in bindings, see e.g. the test test_get_all_variables_overlapping_variables:

        model1 = Gaussian(display_name='model1')

        binding1 = FitBinding(parameter_name='parameter1', model=model1)
        binding2 = FitBinding(parameter_name='parameter2', model=model1)

        parameter_analysis.bindings = [binding1, binding2]

        # THEN
        variables = parameter_analysis.get_all_variables()

        # EXPECT
        assert isinstance(variables, list)
        assert set(variables) == set(model1.get_all_variables())
        assert len(variables) == len(model1.get_all_variables())

(It's more likely to happen with fits to a DiffusionModel)

If there's a smarter way I'm happy to hear it :)

[rozyczko]

This makes testing the plot function very awkward (tests must mock _in_notebook). Consider whether this restriction is necessary. Doesn't plopp support non-notebook figures? Maybe it can be used without the notebook.

[henrikjacobsenfys]

The plopp slicer is only supported in notebooks. I agree it's very awkward, but the alternative is not having the slicer and then the logic gets complicated since I need to figure out how to plot the data then.

[rozyczko]

self._binding is always a list

def _verify_bindings(self, bindings: FitBinding | list[FitBinding] | None) -> list[FitBinding]:

so | None is wrong here.

[rozyczko]

This seems to be unused. Consider removing

[rozyczko]

paramters -> parameters