[tovrstra]

@RobbeBohy @simondc-ugent

While preparing the FAMD project, I realized why the Padé model works well for conductivity and this PR works out the details. It adds a DebyeModel class which complements the LorentzModel for when one expects a minimum in the spectrum at zero frequency. They both work in very similar ways but are typically applied in very different scenarios.

I'll take another good look at everything before merging, but need to leave now. Can you already take a look. I think this can be useful for Computational Physics too.

[simondc-ugent]

Small typo fix:
A valid Debye model has $\hat{C}_\Delta > 0$, which is reached when $p_2 > p_0 * q_2$, not when $p_2 > p_0, q_2$.

The same typo is found in the Lorentz model (line 133), where $p_0 * q_2 > p_2$ should replace $p_0, q_2 > p_2$.

In the source code, it is correctly implemented.

[tovrstra]

Thanks. This is an unfortunate GitHub rendering issue, which does indeed show the comma. When building the documentation with a LaTeX compiler it will show \, as a thin space.

[tovrstra]

It appears to be a known GitHub issue that won't be fixed soon: https://github.com/orgs/community/discussions/17143

[simondc-ugent]

This looks good. I found a small typo in the model.md. Otherwise, the theory and example seem correct. The implementation in the source code is still a bit new to me, so I can't completely vouch for this part. However, I haven't found any issue in the parts I did understand.

[RobbeBohy]

Looks good to me.

[tovrstra]

Thanks for checking