gumbelAnalysis: Gumbel Analysis

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

Description

Estimates and characterizes uncertainty for the distribution of the peak value of a random process using the Gumbel peaks over threshold model.

Usage

gumbelAnalysis(complete_series, length_series, n_min, n_max,
  length_target_series, wplot_filename = NULL, BW = NULL,
  n_mc = 1000, max_dist_hist_filename = NULL, n_boot = 1000,
  ci_level = 0.8)

Arguments

complete_series	(numeric vector) The observed time series that is a realization of the random process.
length_series	(numeric scalar) The length of the time series in units of time (seconds, minutes, hours, etc.).
n_min	(numeric scalar) The minimum number of observations to include in the search for the optimal threshold.
length_target_series	(numeric vector) The distribution of the peak depends on the length of observation, which could be different from the length of the original series. Specify one or more lengths in units of time (seconds, minutes, hours, etc.) for the length of series for which the distribution of the peak is sought. The units of time should be the same as for the argument length_series.
wplot_filename	(NULL or string) If NULL the plot is not created; else, it is saved to "wplot_filename.pdf"; the default is NULL.
BW	(logical) Create the Wplot in black and white (TRUE) or color (FALSE); the default is TRUE.
n_mc	(numeric scalar) The number of Monte Carlo samples from which to construct the distribution of the peak; the defaults is 1000.
max_dist_hist_filename	(NULL or string) If NULL the plot is not created; else, it is saved to "max_dist_hist_filename_length_target_series.pdf"; the default is NULL.
n_boot	(numeric scalar) The number of bootstrap samples for the uncertainty evaluation; the default is 1000.
ci_level	(numeric scalar) The level of the confidence interval for the mean of the peak distribution; the default is 0.8.
m_max	(numeric scalar) The maximum number of observations to include in the search for the optimal threshold.

Details

See the vignette by running vignette(topic = 'potMax_details', package = 'potMax').

Value

Returned invisably, a data frame with four columns. The first column is the estimated mean of the distributio of the peak. The second column is a bootstrap standard error for the mean, and the third and fourth columns provide a percentile bootstrap confidence interval

Examples

## Not run: 
  complete_series <- -jp1tap1715wind270$value
  gumbelAnalysis(complete_series = complete_series,
                 length_series = 100,
                 n_min = 10,
                 n_max = 100,
                 length_target_series = c(100, 167),
                 wplot_filename = './demo/tap1715_dir270_gumbel_wplot',
                 max_dist_hist_filename = './demo/tap1716_dir270_gumbel_max_dist_hist')


## End(Not run)

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