adTestPlot: Threshold diagnostic using the Anderson Darling test

Description Usage Arguments Details See Also Examples

Description

Produces the graphics of the p-values of the Anderson-Darling (AD) test in respect of candidate tresholds.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
adTestPlot(x, ...)

## S3 method for class 'data.frame'
adTestPlot(obj, ...)

## S3 method for class 'matrix'
adTestPlot(obj, ...)

## S3 method for class 'formula'
adTestPlot(form, x, ...)

## S3 method for class 'numeric'
adTestPlot(x, dt = NULL, u, declust = "run", r = 1,
  rlow = 0.75, method.estim = "mle", nboot.estim = 1000,
  method.test = "tab", stat.test = "AD", nboot.test = 1000,
  logit.scale = FALSE, xlab = "Threshold (u)", ylab = "p-value",
  alpha = c(0.05, 0.25), col = "black", lty = 1, lwd = 1,
  col.alpha = "black", lty.alpha = 3, lwd.alpha = 1, cores = NULL,
  display = TRUE, ...)

Arguments

x, dt, form

Sample and time of observation. If a formula is provided, then x must be a data.frame with the corresponding variable.

...

Other graphical parameters

u

Vector of candidate threshold

method.estim, nboot.estim

Parameters for estimation. See fitPot.

method.test, stat.test, nboot.test

Parameters for AD tests. See adTest.

stat.test

Which version of the AD test. One of 'AD','AL' or 'AU', which are respectively the classical AD test and the lower or upper modified AD test.

logit.scale

If true, the p-values are transformed to the negative logit scale. Otherwise the 1 - p-values are plotted.

alpha

Vector indicative references for the p-values.

display

Should the graphics be display. Remark that the p-values are returned by the function.

Details

The function adTestPlot will loop across calls of function fitPot and adTest and will produce a graphics of the P-values in respect of the threshold u. See respective functions for a description of the other function parameters.

See Also

adTest, mrlPlot, dispIndexPlot, which.floodPeaks

Examples

1
2
3
4
## Find list of candidate thresholds
lstu <- seq(500,2500,len = 50)

adTestPlot(flow~date,canadaFlood$daily, u = lstu,declust = 'flood', r = 14)

martindurocher/floodRFA documentation built on June 5, 2019, 8:44 p.m.