Organize your forest plot into collapsible sections using the group parameter. This is ideal for meta-analyses with natural categories like study design, region, or subpopulation.

Basic Grouping

Pass a column name to group to create collapsible sections:

Code 
grouped_data <- data.frame(
  study = c("Study A1", "Study A2", "Study B1", "Study B2", "Study B3"),
  region = c("North", "North", "South", "South", "South"),
  or = c(0.8, 0.9, 1.1, 1.3, 0.95),
  lower = c(0.6, 0.7, 0.85, 1.0, 0.7),
  upper = c(1.1, 1.15, 1.4, 1.7, 1.3)
)

forest_plot(grouped_data,
  point = "or", lower = "lower", upper = "upper",
  label = "study",
  group = "region",
  null_value = 1
)

Click the group headers to collapse or expand sections. Groups are rendered with:

  • Distinct header rows with chevron indicators
  • Automatic indentation of child rows
  • Subtle background shading for visual hierarchy

Hierarchical Grouping

For multi-level organization, pass multiple column names as a vector. webforest creates nested, collapsible hierarchies:

Code 
hierarchical_data <- data.frame(
  study = c(
    "Northeast Trial", "Southeast Trial", "Midwest Trial",
    "Ontario Study", "Quebec Study",
    "England RCT", "Scotland RCT",
    "Bavaria Trial", "Berlin Trial"
  ),
  region = c(
    rep("Americas", 5),
    rep("Europe", 4)
  ),
  country = c(
    rep("USA", 3), rep("Canada", 2),
    rep("UK", 2), rep("Germany", 2)
  ),
  or = c(0.82, 0.91, 0.78, 0.85, 0.88, 0.92, 0.87, 0.79, 0.84),
  lower = c(0.68, 0.75, 0.62, 0.70, 0.72, 0.76, 0.71, 0.64, 0.68),
  upper = c(0.99, 1.10, 0.98, 1.03, 1.07, 1.11, 1.07, 0.98, 1.04)
)

forest_plot(hierarchical_data,
  point = "or", lower = "lower", upper = "upper",
  label = "study",
  group = c("region", "country"),
  null_value = 1, scale = "log",
  theme = web_theme_modern()
)

Each level of the hierarchy:

  • Can be collapsed independently
  • Has progressively deeper indentation
  • Uses subtle visual differentiation (font weight, background)

Customizing Group Headers

Control the visual styling of group headers at each level using theme modifiers:

Code 
forest_plot(hierarchical_data,
  point = "or", lower = "lower", upper = "upper",
  label = "study",
  group = c("region", "country"),
  null_value = 1, scale = "log"
) |>
  set_group_headers(
    level1_font_size = "1.1rem",
    level1_font_weight = 700,
    level1_background = "#e0f2fe",
    level2_font_size = "0.95rem",
    level2_italic = TRUE,
    level2_background = "#f0f9ff",
    indent_per_level = 20
  )

See Themes Reference for all available options.

Starting Collapsed

To have groups start in a collapsed state, you’ll need to use explicit group definitions with web_group():

Code 
forest_plot(grouped_data,
  point = "or", lower = "lower", upper = "upper",
  label = "study",
  group = list(
    web_group("North", collapsed = FALSE),
    web_group("South", collapsed = TRUE)
  ),
  null_value = 1
)

This is useful for large datasets where you want users to expand only the sections they’re interested in.

Grouping vs Row Styling

webforest provides two complementary ways to organize data:

Feature Purpose Best For
group Automatic collapsible sections from column values Categorical data with natural groupings
row_type Manual section headers, summaries, spacers Structured layouts with subtotals

Use group when your data has a grouping column. Use row_type when you need precise control over the layout structure, such as meta-analysis summary rows. See Row Styling for the manual approach.

You can combine both approaches—use group for automatic organization and row_type for summary diamonds within each group.

See Also

  • Row Styling — Manual row types, headers, summaries, and visual styling
  • Themes — Customize group header appearance
  • forest_plot() — Complete argument reference