[WIP] Experiment updates and more flexible cross-validation for Paper revisions

[odunbar]

Purpose

Adjusts examples to fit referee reviews.

• this adds parameter sweeps over n_features in G function example
• this adds tests to sweep over ranks of different nonseparable and separable feature distributions for L63 and UQ for EDMF
• this updates and adds better plotting

Content

Additionally the following src code changes were also made

• movement towards a more concrete cross-validation setup (n_cross_val_sets can be set 0 for no CV, or an integer k to perform k splits with the train-test fraction)
• added the beta / scaled_coeffs penalty back into the loss function
• moved some default values of ensemble size to be 10*n_params capped at 100 inline with our EKP recommendations
• made some unit-tests and online docs compatible with julia v1.11

MISC

• there is a bad Scalar RF test that I will report from line 344
  @test isapprox.(norm(μs - new_outputs), 0, atol = tol_μ)
  Often the problem zeros during optimization (possibly identity issue and should be protected against this)

---

• I have read and checked the items on the review checklist.

[codecov]

Codecov Report

Attention: Patch coverage is 81.90476% with 19 lines in your changes missing coverage. Please review.

Project coverage is 88.85%. Comparing base (acf1051) to head (b636787).
Report is 27 commits behind head on main.

Files with missing lines	Patch %	Lines
src/VectorRandomFeature.jl	67.30%	17 Missing ⚠️
src/ScalarRandomFeature.jl	95.00%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #321      +/-   ##
==========================================
+ Coverage   88.17%   88.85%   +0.67%     
==========================================
  Files           8        7       -1     
  Lines        1184     1274      +90     
==========================================
+ Hits         1044     1132      +88     
- Misses        140      142       +2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.