mrlplot: Mean Residual Life Plot
Description
An empirical mean residual life plot, including confidence intervals, is produced. The mean residual life plot aids the selection of a threshold for the GPD or point process models.
Usage
mrlplot(x, nint = 100, alpha = 0.05, na.action = na.fail, xlab = "Threshold", ...)
Value
A matrix with the mean excess values and their confidence bounds is returned invisibly.
Arguments
- x
numeric vector of data.
- nint
Number of thresholds to use.
- alpha
number giving the 1 - alpha confidence levels to use.
- na.action
function to be called to handle missing values.
- xlab
character string giving the abscissa label to use.
- ...
optional arguments to plot.
Details
The mean excesses are found for each value of a range of thresholds that cover the range of the data (less one at the high end). CIs are also shown based on the normal df for the mean excesses. The goal is to find the lowest threshold such that the graph is linear with increasing thresholds, within uncertainty.
See Coles (2001) sec. 4.3.1 for more information.
References
Coles, S. (2001). An introduction to statistical modeling of extreme values, London, United Kingdom: Springer-Verlag, 208 pp.