[jaspRegression] Change default model prior from beta-binomial to uniform in...

[sisyphus-jasp]

Summary

Fixes: https://github.com/jasp-stats/INTERNAL-jasp/issues/2922

Root cause

Bayesian regression defaults still pointed to betaBinomial (wrapper args + QML default), not uniform. Also no uniform-specific model-size footnote in model comparison output.

What changed

• Set default modelPrior to uniform in:
  • R/regressionlinearbayesianWrapper.R
  • R/regressionlogisticbayesianWrapper.R
  • inst/qml/RegressionLinearBayesian.qml
  • inst/qml/RegressionLogisticBayesian.qml
• Added uniform explanatory footnote in model comparison table code:
  • R/regressionlinearbayesian.R
  • R/regressionlogisticbayesian.R
• Reordered QML radio options so jaspTools QML option parser resolves default to uniform reliably.

Verification

• testAll() result: [ FAIL 0 | WARN 58 | SKIP 1 | PASS 594 ].
• Reproduced with same scenarios for both analyses:
  • default prior now uniform
  • uniform footnote present for default/explicit uniform
  • uniform footnote absent for explicit betaBinomial
  • runs complete successfully.

Caveats

• Existing warning noise unchanged (58 warns), appears baseline-like.
• No test files edited.

Implementation Plan

Root cause

• Bayesian regression defaults are set to betaBinomial in both UI (inst/qml/RegressionLinearBayesian.qml, inst/qml/RegressionLogisticBayesian.qml) and wrappers (R/regressionlinearbayesianWrapper.R, R/regressionlogisticbayesianWrapper.R).
• Model comparison output only adds model-prior explanatory footnotes for wilson and castillo; no explanatory footnote is shown for uniform in R/regressionlinearbayesian.R and R/regressionlogisticbayesian.R.

Proposed changes

• Change default model prior to uniform:
  • In both Bayesian regression QML files, move checked: true from betaBinomial to uniform.
  • In both wrapper functions, change default argument modelPrior = "betaBinomial" to modelPrior = "uniform".
  • Update model-prior description text so the default note is on uniform instead of betaBinomial.
• Add explanatory uniform-prior footnote in both model-comparison table builders:
  • R/regressionlinearbayesian.R in .basregTableModelComparison().
  • R/regressionlogisticbayesian.R in .bayesianLogisticRegTableModelComparison().

Expected test impact

• Existing behavior changes where defaults are used (UI/programmatic default options), so snapshot/expectation differences are possible only in tests relying on defaults.
• Most existing regression tests explicitly set options$modelPrior, so broad regression risk is low.
• New footnote appears only when modelPrior == "uniform"; this may affect tests that inspect that output for uniform settings.

Change default model prior from beta-binomial to uniform in regression analysis and add explanatory footnote about model size benefits

Test Results

Test Run	Result
Baseline (pre-fix)	[ FAIL 1 | WARN 58 | SKIP 1 | PASS 593 ]
Post-fix	[ FAIL 1 | WARN 58 | SKIP 1 | PASS 593 ]
Upstream CI	50926f8 -- CI: passing

Automated Code Review

Approved after 1 review iteration(s).

[codecov-commenter]

Codecov Report

❌ Patch coverage is 62.50000% with 3 lines in your changes missing coverage. Please review.
✅ Project coverage is 75.90%. Comparing base (50926f8) to head (0578cf6).

Files with missing lines	Patch %	Lines
R/regressionlogisticbayesian.R	25.00%	3 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #483      +/-   ##
==========================================
- Coverage   75.91%   75.90%   -0.02%     
==========================================
  Files          21       21              
  Lines        8384     8392       +8     
==========================================
+ Hits         6365     6370       +5     
- Misses       2019     2022       +3     

☔ View full report in Codecov by Sentry.
📢 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.