Some love to model comparison

[Kucharssim]

Model comparison still had some rough edges, so I am taking the liberty to make it smoother.

• Implement a .predict method for the model comparison approximator
• Fixes to the mc_confusion_matrix_plot
• Changed some behavior of the model comparison simulator, which makes it much more useable especially in the case when we want to use mixed batches
• Added a simple example notebook

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

Attention: Patch coverage is 12.50000% with 42 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
...low/approximators/model_comparison_approximator.py	13.63%	19 Missing ⚠️
bayesflow/simulators/model_comparison_simulator.py	13.63%	19 Missing ⚠️
bayesflow/diagnostics/plots/mc_confusion_matrix.py	0.00%	4 Missing ⚠️

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Files with missing lines	Coverage Δ
bayesflow/diagnostics/plots/mc_calibration.py	22.72% <ø> (ø)
bayesflow/diagnostics/plots/mc_confusion_matrix.py	22.50% <0.00%> (-1.19%)	⬇️
...low/approximators/model_comparison_approximator.py	23.89% <13.63%> (-2.48%)	⬇️
bayesflow/simulators/model_comparison_simulator.py	24.07% <13.63%> (-1.51%)	⬇️

[paul-buerkner]

Thank you! Looks good to me. Can you sync this branch with #310 where I made some general interface improvements to the diagnostic functions? Once #310 is merged, I will merge your PR too.

[Kucharssim]

Once #310 is merged, I will merge your PR too.

Sure, I can wait until #310 is merged, it'd easier for me to then get up to date with the dev branch.

[Kucharssim]

it seems the linter again failed for something not in this PR...

[stefanradev93]

That's on me. It's fixed on dev now.