Skip to contents

Plot Hyperedge O-information Estimates

Source: R/plot_hyperedge.R
plot_hyperedge.Rd

Visualize O-information estimates for hyperedges identified during hypergraph construction. Each hyperedge is represented by its O-information value and, optionally, its bootstrap confidence interval.

Usage

plot_hyperedge(
  hg,
  significant_only = FALSE,
  sort = TRUE,
  show_ci = TRUE,
  color_type = TRUE,
  show_edge_label = TRUE,
  point_size = 1.8,
  redundant_color = "#0072B2",
  synergistic_color = "#D55E00"
)

Arguments

hg

A psychHypergraph object produced by build_hypergraph.

significant_only

Logical; if TRUE, only hyperedges passing the significance testing pipeline are displayed.

sort

Logical; if TRUE, hyperedges are sorted by O-information values from largest to smallest.

show_ci

Logical; if TRUE, bootstrap confidence intervals are displayed as horizontal error bars.

color_type

Logical; if TRUE, points are colored according to hyperedge type:

  • Redundant (positive O-information)

  • Synergistic (negative O-information)

show_edge_label

Logical; if TRUE, hyperedge labels are displayed on the y-axis.

point_size

Numeric value controlling point size.

redundant_color

Color for redundant hyperedges (positive O-information).

synergistic_color

Color for synergistic hyperedges (negative O-information).

Value

A ggplot2 object.

The returned object can be further customized using standard ggplot2 syntax.

Details

Positive O-information indicates redundant higher-order interactions, whereas negative O-information indicates synergistic interactions.

The resulting figure resembles a forest plot, allowing users to inspect the magnitude, direction, and uncertainty of hyperedge-level O-information estimates.

Hyperedges are plotted on the y-axis and O-information values on the x-axis. A vertical dashed line at zero separates redundant (\(O > 0\)) and synergistic (\(O < 0\)) interactions.

When bootstrap confidence intervals are available, they are displayed using the BCa intervals stored in the psychHypergraph object.

Hyperedge labels correspond to the variable indices defining each hyperedge. For example:


1-3-5

represents the hyperedge formed by variables 1, 3, and 5.

For large hypergraphs, displaying all hyperedges may result in crowded figures. In such cases, setting significant_only = TRUE or subsetting the hypergraph beforehand is recommended.

See also

build_hypergraph, plot.psychHypergraph

Examples

if (FALSE) { # \dontrun{
library(psych)
data(bfi)

hg <- build_hypergraph(
  data = bfi[,1:25],
  k = 3:5,
  candidate_method = "all",
  cor_method = "pearson",
  parallel = TRUE
)

# Plot all hyperedges
plot_hyperedge(hg)

} # }