Skip to contents

Plot Diagnostics for a Fitted DBN Model

Source: R/dispatch.R
plot.dbn.Rd

Creates model-specific diagnostic plots for a fitted DBN object. The type of plot depends on the model variant:

  • Static: trace plots, posterior histograms, and a network summary of the estimated B matrix.

  • Dynamic: trace plots for variance parameters and, if available, time-varying A/B summaries.

  • Lowrank: trace plots, estimated factor trajectories (\(\alpha_t\)), and the posterior mean node-loading matrix U.

  • HMM: regime probabilities over time, the estimated transition matrix, and MCMC trace plots.

  • Piecewise: trace plots for each regime block.

Usage

# S3 method for class 'dbn'
plot(x, ...)

Arguments

x

A fitted dbn object returned by dbn().

...

Additional arguments passed to model-specific plot functions (e.g., alpha for edge significance in the static model).

Value

A ggplot2 object (or arranged multi-panel plot) is printed and returned invisibly.

See also

dbn(), plot_trace(), check_convergence(), summary_dbn()

Examples

# \donttest{
sim <- simulate_static_dbn(n = 8, time = 10, seed = 6886)
fit <- dbn(sim$Y, model = "static", nscan = 200, burn = 100, verbose = FALSE)
plot(fit)

# }