Add plots for objectives, CV, latent spaces, and contributions

[Jodielidada]

Summary

Adds plot_objective() for recorded optimization histories, plot_cv() for common/separate-penalty heatmaps, plot_source_spaces() for population projection matrices, and plot_contributions() for squared singular-vector entries. Only the CV heatmap needs optional ggplot2.

Plotting behavior and examples

plot_objective() shows all iterations, a numeric selection (iterations beyond the recorded history are ignored), or a window of up to 41 iterations centered on the recorded minimum. README shows an objective plot for the single-source example, and CV output, its heatmap, and a comparison of source choices using dat_multisource. Examples, help pages, and NEWS are updated. plot_cv_profile() from #4 is not carried over, to keep the package easier to maintain.

README also presents the second source comparison added in #5, showing the results when both sources have appreciable noise.

Testing and merge order

Local tests and R CMD check --no-manual --no-build-vignettes pass (macOS, R 4.6.1), and GitHub Actions pass on macOS, Windows, and Ubuntu. Replaces #4 and depends on #5. Please merge #5 with a regular merge commit; if it is squash-merged, I will need to rebase this branch before merging it.

Plot examples

Objective function

The same fit shown over all iterations, iterations 1–200, and a window of 20 iterations on either side of the minimum recorded value; the shaded band shows consecutive objective values, not uncertainty.

Cross-validation

The heatmap compares the candidate penalties, with a black box marking the selected combination.

Source spaces and contributions

The space plots compare the target and two source populations; the contribution plots show squared singular-vector entries for the three components of the fitted target matrix.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 99.18%. Comparing base (a72af46) to head (279c45e).

Additional details and impacted files

@@                       Coverage Diff                       @@
##           multisource-penalties-clean       #6      +/-   ##
===============================================================
+ Coverage                        97.95%   99.18%   +1.23%     
===============================================================
  Files                                5        7       +2     
  Lines                              244      614     +370     
===============================================================
+ Hits                               239      609     +370     
  Misses                               5        5              

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.