riPlot: Plot Relative Information of Estimators

In stamats/rmx: Radius-Minimax Estimators

Plot Relative Information of Estimators

Description

The function riPlot plots the relative information (RI) of estimators. It is only usefull and implemented for models where at least two parameters have to be estimated.

Usage

riPlot(x, ...)

## S3 method for class 'rmx'
riPlot(x, range.alpha = 1e-6, range.n = 501, 
        info.digits = 2, param.digits = 2, 
        ggplot.xlab = "x", ggplot.ylab = NULL, 
        ggplot.ggtitle = NULL,
        point.col = "#0072B5", point.alpha = 0.4, 
        point.length.out = 5, point.range = c(1,7), 
        plot = TRUE, ...)

Arguments

x	object of S3 class rmx.
range.alpha	alpha-quantile used to compute plot range, which is from range.alpha-quantile to 1-range.alpha-quantile of the fitted model.
range.n	number of points in the alpha-quantile range, generated by function seq.
info.digits	number of digits shown for info, which is the absolut/total information assigned to an observation (= length of the IF squared).
param.digits	number of digits used for the estimated parameter values, if default ggplot.ggtitle is used.
ggplot.xlab	labels of x-axes, recycled if length is equal to 1.
ggplot.ylab	if NULL default labels of y-axes are generated. If it is specified, it should have the same length as the number of parameters. If length is equal to 1, it will be recycled.
ggplot.ggtitle	if NULL default plot titles for each component of the IF are generated. If it is specified, it should have the same length as the number of parameters. If length is equal to 1, it will be recycled.
point.col	single color used for colouring the data points.
point.alpha	alpha used for color shading.
point.length.out	number of point sizes used to visualize the absolut/total information assigned to a data point; see argument breaks of function scale_size.
point.range	a numeric vector of length 2 specifing the minimum and maximum size of the data points; see argument range of function scale_size
plot	if FALSE, the plotting is suppressed. Can be used to modify the invisibly returned plotting object.
...	further arguments passed through.

Details

The function is inspired by the plot-methods and function PlotIC of the RobASt-family of packages.

In case of optimally-robust RMX estimators computed with function rmx (S3 class rmx), the relative information is plotted which corresponds to the absolute value of the the respective component/coordinate of the influence function devided the absolute information. The size of the data points represents the absolute information.

Value

Invisible object of class ggplot or a list of ggplot objects.

Author(s)

Matthias Kohl [e-mail]

References

Kohl, M. (2005) Numerical Contributions to the Asymptotic Theory of Robustness. Bayreuth: Dissertation.

Rieder, H. (1994) Robust Asymptotic Statistics. New York: Springer.

Rieder, H., Kohl, M. and Ruckdeschel, P. (2008) The Costs of not Knowing the Radius. Statistical Methods and Applications 17(1) 13-40. Extended version: http://r-kurs.de/RRlong.pdf

M. Kohl, P. Ruckdeschel, and H. Rieder (2010). Infinitesimally Robust Estimation in General Smoothly Parametrized Models. Statistical Methods and Application, 19(3):333-354.

See Also

rmx, optIF

Examples

ind <- rbinom(100, size=1, prob=0.05) 
x <- rnorm(100, mean=ind*3, sd=(1-ind) + ind*9)
res <- rmx(x, eps.lower = 0.01, eps.upper = 0.1)
riPlot(res)

## plot-method
plot(res, which = 6)

## setting and passing argument
riPlot(res, ggplot.xlab = "data", ggplot.ylab = "relative information",
       ggplot.ggtitle = c("Rel. information of location part",
                          "Rel. information of scale part"))
plot(res, which = 6, 
     control = list(riPlot = list(ggplot.xlab = "data", 
                                  ggplot.ylab = "relative information",
                                  ggplot.ggtitle = c("Rel. information of location part",
                                                     "Rel. information of scale part"))))

stamats/rmx documentation built on Sept. 29, 2023, 7:13 p.m.