Trace plots for MCMC diagnostics

Creates trace plots to assess MCMC convergence. Supports plotting either the graph similarity parameter (theta) or the number of edges per group across saved iterations.

Usage

plot_trace(fit, what = c("theta", "edges"))

Arguments

fit

A multiggm_fit object returned by multiggm_mcmc.

what

Character; what to plot:

"theta"

Graph similarity parameters \(\theta_{km}\) for each pair of groups. Requires K >= 2.

"edges"

Number of edges per group at each saved iteration. Color-coded by group.

Value

A ggplot object.

See also

plot.multiggm_fit(), edge_counts()

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_trace(fit, "theta")

plot_trace(fit, "edges")