Viz 02: Harmonise 1D distribution plots (histogram + KDE)

[Doresic]

It makes sense to make parameter_hist and sampling_1d_marginals (the two 1D distribution plotters) to behave the same: same bounds handling, same axis labels, same legend layout, same argument vocabulary, as they are very similar plotters.

Main changes:

• both functions now take show_bounds: bool = True — when true, the plot frames to the parameter bounds and draws bound lines; when false, it frames tightly to the data. xlim is set explicitly from bounds + actual data, so the KDE tail extension never dominates the visible range
• scale-aware x-labels (log parameters get a log-scale axis label)
• per-panel legend with KDE / Samples / Bounds entries
• rug uses the darker dash_color (matching the bar-edge family) so individual data points stay readable on the lighter histogram
• new shared helper plot_density_panel in misc.py doing the histogram / KDE / rug / bounds rendering — takes a resolved style dict
• harmonised the two siblings' arguments: same shared-kwarg order (plot_type, bins, bw_method, show_bounds, title, size, ax/axes, style_kwargs), and sampling_1d_marginals gained bins (was hardcoded). Tidied up some confusing docstrings along the way (stepsize is a thinning factor; start_indices subsets starts while parameter_indices subsets parameters)
• new apply_style() in _style.py — a small opt-in rcParams preset (top/right spines off, framed legends, constrained layout). General so later PRs in the series just call it
• new generic style keys: LINE_* for any line (KDE, simulation, …), DASH_* for any tick marker (rug, CI endpoints, …). The naming principle is documented in _style.py so future PRs follow it
• parameter_hist gains bw_method and title; its legacy color kwarg is deprecated — use style_kwargs={"rectangle_color": ..., "line_color": ..., "dash_color": ...} instead
• process_deprecated_kwarg in misc.py gains a removal mode (canonical_name=None) and a note= parameter for custom messages — used here for the color deprecation; PR 09b will reuse it
• GRID_SIZE_PER_COL / GRID_SIZE_PER_ROW in _style.py wire into get_axes_array's size=None default so every multi-panel plotter inherits consistent panel sizing

This is the second in the small series of visualisation PRs. Before/after PNGs below.

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

❌ Patch coverage is 90.19608% with 15 lines in your changes missing coverage. Please review.
✅ Project coverage is 84.38%. Comparing base (450b571) to head (0ded28c).

Files with missing lines	Patch %	Lines
pypesto/visualize/misc.py	76.19%	10 Missing ⚠️
pypesto/visualize/_style.py	91.30%	4 Missing ⚠️
pypesto/visualize/sampling.py	97.14%	1 Missing ⚠️
❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@                    Coverage Diff                     @@
##           viz-01-cluster-palette    #1718      +/-   ##
==========================================================
+ Coverage                   84.33%   84.38%   +0.05%     
==========================================================
  Files                         165      165              
  Lines                       14796    14926     +130     
==========================================================
+ Hits                        12478    12596     +118     
- Misses                       2318     2330      +12     

☔ 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.

[Doresic]

Here are the plot comparisons
before and after
Histograms before:

Histograms after:

1d-sampling-marginals before:

1d-sampling-marginals after no bounds:

1d-sampling-marginals after with bounds: