Issue 54: make colorblind friendly

[kaitejohnson]

• Add shapes as an additional distinguisher between models and makes legends

Summary by CodeRabbit

• New Features

  • Enhanced model differentiation in plots through visual shape indicators alongside colour coding.
  • Configurable legend display for improved plot flexibility.

• Improvements

  • Refined legend visibility controls across visualisations.
  • Improved visual consistency and aesthetic refinements in plot presentations.

_{✏️ Tip: You can customize this high-level summary in your review settings.}

[coderabbitai]

Walkthrough

This pull request introduces shape aesthetics for models in plotting functions, adds a new add_shape parameter to control shape legend display, and expands legend handling across plotting targets by introducing remove_legend parameters to control legend visibility in various plots.

Changes

Cohort / File(s)	Summary
Overall scores plotting function R/fig_overall_scores.R	Adds shape mapping for models alongside existing colour mappings. Introduces add_shape parameter (default: FALSE) to optionally display shape legend. Updates get_plot_overall() function signature. Adjusts legend guides and layout, removes redundant axis labels, and refactors legend handling when remove_legend is TRUE.
Plotting style configuration R/plotting_style.R	Introduces model_shapes mapping for model names to numeric shape codes. Includes model_shapes in plot_comp_list output for shape control.
Plot targets configuration targets/fig_overall_targets.R	Adds remove_legend parameter to multiple plotting targets: CA-related bar charts and relative skill plots, and by-location relative skill plot. Sets visibility to TRUE or FALSE as appropriate for each target.

Possibly related PRs

• epiforecasts/evalvariantnowcasthub#20: Touches the same plotting code in R/fig_overall_scores.R and R/plotting_style.R, including modifications to get_plot_overall and plot styling components.

Pre-merge checks and finishing touches

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title references Issue 54 and indicates the objective (colorblind friendly) but the actual changeset focuses on adding shape aesthetics to distinguish models, which is a technical implementation detail not explicitly mentioned in the title.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

✨ Finishing touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch 54-colorblind-friendly

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 0

Caution

Some comments are outside the diff and can’t be posted inline due to platform limitations.

⚠️ Outside diff range comments (1)

R/fig_overall_scores.R (1)

603-637: Missing scale_shape_manual in absolute scores block.

Whilst the shape aesthetic is mapped to model at line 621 and the shape legend is configured in guides, the scale_shape_manual call is missing. This means ggplot2 will use default shapes rather than the custom model_shapes defined in plot_components(), defeating the accessibility purpose of this PR.

In contrast, the relative skill block (lines 560-596) correctly includes scale_shape_manual at line 584.

🔎 Proposed fix

       )
     ) +
     scale_color_manual(values = plot_components_list$model_colors) +
+    scale_shape_manual(values = plot_components_list$model_shapes) +
     geom_vline(xintercept = 0, linetype = "dashed", color = "gray50") +
     get_plot_theme() +

🧹 Nitpick comments (1)

R/fig_overall_scores.R (1)

500-515: Clarify the interaction between remove_legend and add_shape parameters.

The current logic allows add_shape = TRUE to override remove_legend = TRUE for the shape legend specifically. This creates an asymmetric API where:

• If remove_legend = TRUE and add_shape = TRUE: shape legend shows, but colour/fill legends are hidden
• If remove_legend = TRUE and add_shape = FALSE: all legends are hidden

Whilst this may be intentional, consider documenting this behaviour explicitly in the function documentation to prevent confusion.

💡 Suggested documentation enhancement

 #' @param remove_legend Boolean indicating whether to keep legend, default
 #'   is TRUE.
 #' @param add_shape Boolean indicating whether to add the shape legend,
-#'  default is FALSE.
+#'  default is FALSE. Note: when TRUE, this overrides remove_legend for
+#'  the shape aesthetic only.

📜 Review details

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 14fed30 and b38785c.

📒 Files selected for processing (3)

• R/fig_overall_scores.R
• R/plotting_style.R
• targets/fig_overall_targets.R

🔇 Additional comments (5)

R/plotting_style.R (1)

60-67: Excellent addition of shape aesthetics for accessibility.

The shape mappings are well-chosen and use standard ggplot2 shape codes that will provide clear visual distinction between models for colorblind users.

targets/fig_overall_targets.R (1)

105-105: Legend visibility controls are well-coordinated.

The selective legend removal for CA plots and by-location plots aligns well with the multi-panel figure layout strategy where legends are collected at the top level.

Also applies to: 113-114, 122-123, 222-223

R/fig_overall_scores.R (3)

74-111: Shape aesthetics correctly integrated for accessibility.

The implementation properly adds shape mapping alongside colour, includes the appropriate scale_shape_manual, and coordinates legend guides for both aesthetics.

---

197-227: Shape aesthetics consistently applied across both plot types.

Both the relative skill and absolute score paths correctly implement shape mapping with proper scales and legend configurations.

Also applies to: 247-296

---

674-675: Consistent bar chart styling applied.

The use of fixed black fill for sequence count bar charts provides clear, simple visualisation without redundant legend items.

Also applies to: 700-701, 728-729