Skip to contents

For an optimized fit, draws the log posterior and the gradient norm against iteration, from the record the engine kept while fitting. For a sampled fit, draws each chain's draws of every population mean and standard deviation, on the scale summary() reports them on, so chains that have not settled on one distribution can be seen.

Usage

ctTracePlot(fit, which = NULL, plot = TRUE)

Arguments

fit

A fit from ctFit with the Julia backend.

which

Which panels to draw: trace columns ('objective', 'gradient_norm', 'inner_converged') for an optimized fit, parameter names for a sampled one. Defaults to all of them.

plot

If FALSE, return the data rather than a plot: the trace for an optimized fit, one row per draw and parameter for a sampled one.

Value

A ggplot, or with plot = FALSE a data frame.

Details

The Julia backend records the optimisation trace during the fit and returns it as fit$optim$trace. For output while a fit is still running, see optimcontrol$callback in ctFit. The chains of a sampled stan fit are drawn by plot(fit, types = 'trace').

Examples

# \donttest{
# Needs a working julia backend, so the example is inert where Julia is
# absent -- including on CRAN, whose check machines have none.
if (isTRUE(ctJuliaStatus()$available)) {
  model <- ctModel(type = 'ct', manifestNames = 'Y1', latentNames = 'eta1',
    LAMBDA = matrix(1))
  model$pars$indvarying <- FALSE
  fit <- ctFit(ctstantestdat, model, backend = 'julia', cores = 1)
  ctTracePlot(fit)
  # Julia keeps running between calls, holding what it has compiled;
  # this ends it.
  ctJuliaKill("session")
}
# }