plot-LagEstimator {quantspec}

Plot the values of a LagEstimator.

Description

Creates a K x K plot displaying all levels combinations from the argument levels. In each of the subplots either the real part (on and below the diagonal; i. e., \tau_1 \leq \tau_2) or the imaginary parts (above the diagonal; i. e., \tau_1 > \tau_2) of the lag-window estimator, for the combination of levels \tau_1 and \tau_2 denoted on the left and bottom margin of the plot are displayed.

Usage

## S4 method for signature 'LagEstimator,ANY'
plot(
  x,
  ptw.CIs = 0.1,
  ratio = 3/2,
  widthlab = lcm(1),
  xlab = expression(omega/2 * pi),
  ylab = NULL,
  type.scaling = c("individual", "real-imaginary", "all"),
  frequencies = x@frequencies,
  type.CIs = c("naive.sd"),
  levels = intersect(x@levels[[1]], x@levels[[2]])
)

Arguments

x

The LagEstimator object to plot

ptw.CIs

the confidence level for the confidence intervals to be displayed; must be a number from [0,1]; if null, then no confidence intervals will be plotted.

ratio

quotient of width over height of the subplots; use this parameter to produce landscape or portrait shaped plots.

widthlab

width for the labels (left and bottom); default is lcm(1), cf. layout.

xlab

label that will be shown on the bottom of the plots; can be an expression (for formulas), characters or NULL to force omission (to save space).

ylab

label that will be shown on the left side of the plots; can be an expression (for formulas), characters or NULL to force omission (to save space).

type.scaling

a method for scaling of the subplots; currently there are three options: "individual" will scale each of the K^2 subplots to minimum and maximum of the values in that plot, "real-imaginary" will scale each of the subplots displaying real parts and each of the subplots displaying imaginary parts to the minimum and maximum of the values display in these subportion of plots. The option "all" will scale the subplots to the minimum and maximum in all of the subplots.

frequencies

a set of frequencies for which the values are to be plotted.

type.CIs

indicates the method to be used for determining the confidence intervals; the methods available are those provided by getPointwiseCIs-LagEstimator.

levels

a set of levels for which the values are to be plotted.

Value

Returns the plot described in the Description section.

See Birr et al. (2015)

References

Birr, S., Volgushev, S., Kley, T., Dette, H. & Hallin, M. (2015). Quantile Spectral Analysis for Locally Stationary Time Series. http://arxiv.org/abs/1404.4605.