Add method to fit Huld model

[cwhanse]

• [ ] Closes #xxxx
• I am familiar with the contributing guidelines
• Tests added
• Updates entries in docs/sphinx/source/reference for API changes.
• Adds description and name entries in the appropriate "what's new" file in docs/sphinx/source/whatsnew for all changes. Includes link to the GitHub Issue with :issue:`num` or this Pull Request with :pull:`num`. Includes contributor name and/or GitHub username (link with :ghuser:`user`).
• New code is fully documented. Includes numpydoc compliant docstrings, examples, and comments where necessary.
• Pull request is nearly complete and ready for detailed review.
• Maintainer: Appropriate GitHub Labels (including remote-data) and Milestone are assigned to the Pull Request and linked Issue.

[cwhanse]

Test failures are unrelated. Coverage decrease is related to importing statsmodels.

I think I have addressed the maintainer's comments and concerns.

[kandersolar]

LGTM

[adriesse]

We have one thing that is not fully resolved:

AFAIK Huld does not fully describe his fitting method, so a discussion may be needed to assess whether this requires a new publication. See #1898.

I would suggest this PR join the queue, and perhaps more discussion will ensue. The upcoming PVPMC is also a great opportunity to discuss and perhaps develop a policy.

[RDaxini]

We have one thing that is not fully resolved:

AFAIK Huld does not fully describe his fitting method, so a discussion may be needed to assess whether this requires a new publication. See #1898.

I would suggest this PR join the queue, and perhaps more discussion will ensue. The upcoming PVPMC is also a great opportunity to discuss and perhaps develop a policy.

Has anyone tried contacting Thomas Huld or any of the other authors to ask what fitting method was used?

[cwhanse]

Has anyone tried contacting Thomas Huld or any of the other authors to ask what fitting method was used?

I have reached out to several of the authors and have gotten no response.