plotDiagnostics: Plot results from diagnostics collected during model fitting

In owl: Generalized Linear Models Regularized with the Sorted L1-Norm

Description

This function plots various diagnostics collected during the model fitting resulting from a call to owl() provided that diagnostics = TRUE.

Usage

plotDiagnostics(
  object,
  ind = max(object$diagnostics$penalty),
  xvar = c("time", "iteration"),
  yvar = c("objectives", "infeasibility"),
  ...
)

Arguments

object	an object of class "Owl".
ind	either "last"
xvar	what to place on the x axis. iteration plots each iteration, time plots the wall-clock time.
yvar	what to place on the y axis. objectives returns the primal and dual objectives whereas infeasibility returns the infeasibility metric.
...	other arguments that will be used to modify the call to lattice::xyplot()

Value

An object of class "trellis", which, unless stored in a variable, will be plotted when its default print() method is called.

Examples

x <- owl(abalone$x, abalone$y, sigma = 2, diagnostics = TRUE)
plotDiagnostics(x)

owl documentation built on Feb. 11, 2020, 5:09 p.m.