ENH: Add masking support for many algorithms

[derb12]

Description

Mostly addresses issue #72. Adds a new mask property to Baseline and Baseline2D to allow omitting regions from fitting, and adds test cases to ensure correct behavior. This PR adds mask support for all weighted algorithms (polynomial, Whittaker, and p-spline methods) except loess, which will be covered in a separate PR, plus a few optimizers. There are maybe a few other methods that could get mask support without much effort, but they can be added in a later follow up.

Type of Pull Request

• Bug Fix
• New Feature
• Miscellaneous Changes (refactor, code improvements, etc.)
• Documentation or Example Programs

Pull Request Checklist

• New code and/or documentation is valid for use with the BSD 3-clause license.
• New code is fully documented with docstrings that follow Numpy style,
  if applicable.
• New code follows PEP 8 standards as closely as possible, if applicable.
• Added/updated tests and ensured they pass locally, if applicable.
• Verified that documentation builds locally, if applicable.

[derb12]

test_whittaker.py::TestAsPLS::test_pentadiagonal_solver is failing again. It looks like the solution is a little unstable, with 3.10, 3.11, 3.12, and nightly failing with rtol 0.00016443 and atol 0.082209809, while 3.13 and 3.14 fail with rtol 0.00016311 and atol 0.08167898. For now I'll just bump the rtol again, but might want to change the data it's fitting so the fit's more stable in the future.

test_whittaker.py::TestDerpsalsa::test_masking_fit_all only fails on the nightly; just need a small rtol bump.

[derb12]

rtd failure is from the masking example, which will be removed/rewritten on the docs branch, so not worth triggering CI again to fix.