est_tlmcov {TLMoments}

Estimate the covariance matrix of TL-moments estimations

Description

Internal function. Use est_cov. Description not done yet.

Usage

est_tlmcov(
  x,
  leftrim = 0L,
  rightrim = 0L,
  order = 1:3,
  distr = NULL,
  lambda.cov = TRUE,
  ratio.cov = TRUE,
  ...
)

## S3 method for class 'numeric'
est_tlmcov(
  x,
  leftrim = 0L,
  rightrim = 0L,
  order = 1:3,
  distr = NULL,
  lambda.cov = TRUE,
  ratio.cov = TRUE,
  ...
)

## S3 method for class 'matrix'
est_tlmcov(
  x,
  leftrim = 0L,
  rightrim = 0L,
  order = 1:3,
  distr = NULL,
  lambda.cov = TRUE,
  ratio.cov = TRUE,
  reg.weights = NULL,
  ...
)

## S3 method for class 'TLMoments'
est_tlmcov(
  x,
  leftrim = attr(x, "leftrim"),
  rightrim = attr(x, "rightrim"),
  order = attr(x, "order"),
  distr = NULL,
  lambda.cov = TRUE,
  ratio.cov = TRUE,
  set.n = NA,
  ...
)

Arguments

x

numeric vector or matrix containing data OR an object of TLMoments.

leftrim, rightrim

integer indicating lower and upper trimming parameters, have to be non-negative integers.

order

numeric vector giving the orders that are returned (default is first three L-moments).

distr

character of length 1 giving the distribution if parametric assumption should be used.

lambda.cov

boolean, if TRUE (default) TL-moment estimation covariance matrix is calculated.

ratio.cov

boolean, if TRUE (default) TL-moment-ratio estimation covariance matrix is calculated.

...

additional arguments.

reg.weights

numeric vector of weights for regionalized TLMoments.

set.n

hypothetical data length n if theoretical values are given.

Value

a list of numeric matrices (if lambda.cov and ratio.cov are TRUE (default)), or a single matrix.

Examples

### Numeric vectors
x <- rgev(500, loc = 10, scale = 5, shape = .1)

est_tlmcov(x)
est_tlmcov(x, order = 2:3)
est_tlmcov(x, rightrim = 1, order = 4:5)
# cov(t(replicate(10000,
#   TLMoments(rgev(500, loc = 10, scale = 5, shape = .1))$lambdas)
# ))
# cov(t(replicate(10000,
#   TLMoments(rgev(500, loc = 10, scale = 5, shape = .1))$ratios)
# ))

est_tlmcov(x, ratio.cov = FALSE)
est_tlmcov(x, lambda.cov = FALSE)

est_tlmcov(x, distr = "gev")

est_tlmcov(x, leftrim = 0, rightrim = 1)
# cov(t(replicate(10000,
#  TLMoments(rgev(500, loc = 10, scale = 5, shape = .1), 0, 1, 3)$lambdas
# )))
# cov(t(replicate(10000,
#  TLMoments(rgev(500, loc = 10, scale = 5, shape = .1), 0, 1, 3)$ratios
# )))

### Numeric matrices
x <- matrix(rgev(600), nc = 3)

est_tlmcov(x)
est_tlmcov(x, order = 3:4)
# cov(t(replicate(10000,
#   as.vector(TLMoments(matrix(rgev(600), nc = 3))$lambdas[3:4, ])
# )))
# cov(t(replicate(10000,
#   as.vector(TLMoments(matrix(rgev(600), nc = 3))$ratios[3:4, ])
# )))

est_tlmcov(x, ratio.cov = FALSE)
est_tlmcov(x, lambda.cov = FALSE)

TLMoments:::est_tlmcov(x, order = 2:3, distr = "gev")
# cov(t(replicate(10000,
#   as.vector(TLMoments(matrix(rgev(600), nc = 3))$lambdas[2:3, ])
# )))
# cov(t(replicate(10000,
#   as.vector(TLMoments(matrix(rgev(600), nc = 3))$ratios[2:3, ])
# )))

### TLMoments-object (theoretical calculation)
tlm <- TLMoments(as.parameters(loc = 10, scale = 5, shape = .1, distr = "gev"), 0, 1)
est_tlmcov(tlm, distr = "gev", set.n = 100)
est_tlmcov(tlm, distr = "gev", set.n = 100, ratio.cov = FALSE)
est_tlmcov(tlm, distr = "gev", set.n = 100, lambda.cov = FALSE)