feat: limit parameter variations in rankings to stay within parameter bounds

[alexander-held]

As identified by @malin-horstmann, we can run into scenarios where the ranking will attempt to vary parameters beyond their bounds. This can happen with parameters that have asymmetric uncertainties when only a symmetric estimate was done from the Hessian (for shapesys in particular this is likely to happen).

This PR limits the variations to stick within bounds. Things to think about before finishing this:

• raise a warning? (seems useful, but hard to miss in verbose ranking output)
• allow for a way to propagate MINOS uncertainties through instead?
• visually flag such cases in the plot?

The CI fail is currently expected (bounds kwarg changes).

* force parameter variations in ranking to stay within parameter bounds