confint: Generic Function to Compute (Profile) Confidence Intervals In POT: Generalized Pareto Distribution and Peaks Over Threshold

Description

Compute (profile) confidence intervals for the scale, shape GPD parameters and also for GPD quantiles.

Usage

 ```1 2 3``` ```## S3 method for class 'uvpot' confint(object, parm, level = 0.95, ..., range, prob, prof = TRUE) ```

Arguments

 `object` `R` object given by function `fitgpd`. `parm` Charater string specifies for which variable confidence intervals are computed. One of `"quant"`, `"scale"` or `"shape"`. `level` Numeric. The confidence level. `...` Optional parameters. See details. `range` Vector of dimension two. It gives the lower and upper bound on which the profile likelihood is performed. Only required when "prof = TRUE". `prob` The probability of non exceedance. `prof` Logical. If `TRUE` (the default), profile confidence intervals are computed. Otherwise, it is Fisher ones.

Details

Additional options can be passed using "..." in the function call. Possibilites are related to the specific functions: `link{gpd.fiscale}`, `link{gpd.fishape}`, `link{gpd.firl}`, `link{gpd.pfscale}`, `link{gpd.pfshape}`, `link{gpd.pfrl}`.

Value

Returns a vector of the lower and upper bound for the (profile) confidence interval. Moreover, a graphic of the profile likelihood function is displayed when `prof = TRUE`.

Author(s)

Mathieu Ribatet

`link{gpd.fiscale}`, `link{gpd.fishape}`, `link{gpd.firl}`, `link{gpd.pfscale}`, `link{gpd.pfshape}` and `link{gpd.pfrl}`

Examples

 ```1 2 3 4``` ```x <- rgpd(100, 0, 1, 0.25) mle <- fitgpd(x, 0) confint(mle, prob = 0.2) confint(mle, parm = "shape") ```

Example output

```If there is some troubles try to put vert.lines = FALSE or change
the range...
conf.inf  conf.sup
0.2062064 0.3685047
If there is some troubles try to put vert.lines = FALSE or change
the range...
conf.inf   conf.sup
-0.1097566         NA
Warning message:
In min(index.neg2) : no non-missing arguments to min; returning Inf
```

POT documentation built on May 2, 2019, 7:30 a.m.