gumbelNYearUncert.gumbel_multi_fit: Uncertainty in Return Values for the Gumbel Model Using Many...

In usnistgov/potMax: Empirical Calculation of the Distribution of the Maximum from a Peaks Over Threshold Process

Description

The function produces a bootstrap sample of return values using the Gumbel like 2D extremal Poisson process with many possible thresholds

Usage

## S3 method for class 'gumbel_multi_fit'
gumbelNYearUncert(x, declust_obs, N, n_boot,
  progress_tf = TRUE)

Arguments

x	An S3 object of type gumbel_multi_fit.
declust_obs	(numeric vector) The observed data used by gumbelMultiFit. This will very often be the $declustered_series element of an S3 object of class declustered_series
N	(numeric scalar) The N in N-year return value. This is a bit of a misnomer since the unit of time does not have to be years. The function can calculate N-second, N-minute, N-hour, etc. return values as well. In fact, the unit of time is the same unit of time passed in for the lt argument of gumbelMLE. Naming the function for the unit of time year is simply due to my past experince with calculating return values on the time scale of years.
n_boot	(numeric scalar) The number of bootstrap samples
progress_tf	(logical scalar) Display a progress bar if TRUE, else not.

Details

Repeatedly solves the equation

\int_{y_N}^∞\int_0^1λ(t, y)dtdy = \frac{1}{N}

where λ(t, y) is given in the documentation for gumbelMLE for many thresholds and n_boot bootstrap replicates of the unthresholded data inputted togumbelMultiFit.

Value

An S3 object of class gumbel_N_year_val_uncert_multi_thresh which is a list of length n_boot of S3 objects of class gumbel_N_year_val_multi_thresh.

Examples

## Not run: 

complete_series <- -jp1tap1715wind270$value

declustered_obs <- decluster(complete_series)

gumbel_multi_fit <- gumbelMultiFit(x = declustered_obs,
                                   lt = 100,
                                   n_min = 10,
                                   n_max = 50,
                                   weight_scale = 5)

500_second_val_uncert <- gumbelNYearUncert(x = gumbel_multi_fit,
                                           declust_obs = declustered_obs$declustered_series,
                                           N = 500,
                                           n_boot = 200)

## End(Not run)

usnistgov/potMax documentation built on May 3, 2019, 2:38 p.m.