sub_sample_fit: Fitting the LTME or EV Trawl model, potentially on block...

In valcourgeau/gammaextremes: Univariate Extreme Values with Trawl Processes

Description

Fitting the LTME or EV Trawl model, potentially on block subsamples.

Usage

sub_sample_fit(
  data,
  sample_length,
  depth,
  method,
  mode,
  type,
  bounds,
  trials,
  parallel = F,
  seed = NULL
)

Arguments

data	Data to fit.
sample_length	Length of block subsamples.
depth	Depth to fit.
method	Selecting "GMM" or "PL".
mode	Either one-stage/full ("full") or two-stage ("two-stage").
type	Trawl type (e.g. "exp", "sum_exp", etc.)
bounds	Parameters bounds. "config" uses trawl cfg, "multiplier" uses scaled initial guess.
trials	Number of blocks on which to perform the inference.
parallel	Parallel or not, defaults to FALSE.
seed	Seed for block selection, not used by default NULL.

Value

Vector of fitted parameters.

Examples

n <- 2000
test_column <- 2
data <- pollution_data[seq_len(n), test_column]
depth <- 5
ev_trawl_fit(data, depth, method = "GMM")

valcourgeau/gammaextremes documentation built on Sept. 9, 2021, 5:42 a.m.