Skip to contents

Plot method for multiggm_fit objects

plot.multiggm_fit.Rd

Produces diagnostic and summary plots for a fitted multi-GGM model.

Usage

# S3 method for class 'multiggm_fit'
plot(
  x,
  type = c("trace_theta", "trace_edges", "pip", "network"),
  pip_threshold = 0.5,
  ...
)

Arguments

x

A multiggm_fit object returned by multiggm_mcmc.

type

Character string specifying the plot type:

"trace_theta"

Trace plots of the graph similarity parameter \(\theta_{km}\) across post-burn-in iterations. Values > 0 indicate the model is borrowing strength between groups. Requires K >= 2.

"trace_edges"

Trace of the number of included edges per group across iterations. Useful for assessing convergence and stationarity of model complexity.

"pip"

Heatmap of posterior inclusion probabilities (PIP) for each edge, faceted by group. Uses viridis color scale.

"network"

Network visualization of the estimated graph at the given PIP threshold, using igraph. One plot per group.

pip_threshold

Numeric; PIP threshold for edge selection in network plots. Default 0.5. Ignored for other plot types.

...

Additional arguments (currently ignored).

Value

For "trace_theta", "trace_edges", and "pip": a ggplot object (can be further customized). For "network": invisible NULL (plots are drawn as a side effect using igraph).

See also

plot_trace(), plot_pip_heatmap(), plot_network(), plot_roc(), plot_recovery()

Examples

sim <- simulate_multiggm(K = 2, p = 8, n = 80, seed = 1)
fit <- multiggm_mcmc(data_list = sim$data_list, burnin = 200, nsave = 100)
plot(fit, type = "trace_edges")

plot(fit, type = "pip")