arima.contour.plot: ARIMA Model Likelihood Contour Plot

View source: R/arima.contour.plot.R

arima.contour.plotR Documentation

ARIMA Model Likelihood Contour Plot

Description

Provides a contour plot of the likelihood or two-dimensional profile likelihood for a specified ARIMA model and time series data object.

Usage

arima.contour.plot(
  data.tsd,
  model,
  param1 = list(1, seq(-0.99, 0.99, length = 8)),
  param2 = list(2, seq(-0.99, 0.99, length = 8)),
  relative = FALSE,
  verbose = FALSE,
  my.title,
  persp.plot = FALSE,
  ...
)

Arguments

data.tsd

Time series data object.

model

Time series model.

param1

List containing the integer parameter number of the first parameter and sequence of values at which the likelihood will be evaluated.

param2

List containing the integer parameter number of the second parameter and sequence of values at which the likelihood will be evaluated.

relative

If true, the relative likelihood is plotted instead of the log likelihood (default is FALSE)

verbose

Prints model coefficient values for each evaluation point; useful to debug estimation problems when doing a profile plot (when the number of parameters is greater than 2).

my.title

A title (or an empty character string) to replace the default title.

persp.plot

If TRUE a perspeciive plot (instead of a contour plot) will be provided.

...

Allows sending down extra arguments to the contour or persp functions (such as graphs parameters).

Value

Invisibly returns the information used to make the contour plot.

See Also

arima.likelihood.plot.

Examples

#ar1=2 model
arima.contour.plot(device.inventory.tsd, model = model.pdq(p = 2),
         list(1, seq(-1.99, 1.99, length = 50)),
         list(2, seq(-.99, .99, length = 50)))
#ar1=2 model--zoom in
## Not run: arima.contour.plot(device.inventory.tsd, model = model.pdq(p = 2),
         list(1, seq(1.35, 1.394, length = 40)),
         list(2, seq(-0.408, -0.390, length = 40)), my.title='')
## End(Not run)


wqmeeker/RTseries documentation built on Dec. 31, 2022, 10 a.m.