DispIndex: Dispersion index
In martindurocher/floodStat: FloodNet Regional Frequency Analysis

Description Usage Arguments Details Reference See Also Examples

Returns the dispersion index (or variance to mean ratio) to verify the hypothesis of a Homogenous Poisson process.

1 2	DispIndex(obj, k = NULL, ci = FALSE, alpha = 0.05, method = "asymptotic", nsim = 10000)

`obj`	An output of `FitPot`.
`k`	Number of classes. Default value chosen to ensure at least 5 classes and aims at approximately an average 5 events per classe.
`ci`	Should confident interval and p-value be returns.
`alpha`	Significance level.
`method`	Method for computing the confident interval and the p-value. Either 'asymptotic' for chi-squared approximation or 'boot' for bootstrap.
`nsim`	Number of bootstrap samples.

The function PlotDispIndex will loop across calls of function FitPot and DispIndex and will produce a graphics of the Dispersion index in respect of the threshold u. See respective functions for a description of the function parameters. The points in the plot are returns as a data.frame and parameter display can be set to FALSE to prevent the display of the graphic.

Lang M, Ouarda TBMJ, Bobée B. (1999) Towards operational guidelines for over-threshold modeling. Journal of Hydrology. Dec 6;225(3):103–17.

PlotMrl

fit <- FitPot(flow~date,flowStJohn, u = 1000,
              declust = 'wrc', r = 14)

DispIndex(fit, ci = TRUE)

PlotDispIndex(flow~date,flowStJohn,
              u = seq(500,2000, len = 50),
              declust = 'wrc', r = 14)