elife: Excess lifetime distributions

In longevity: Statistical Methods for the Analysis of Excess Lifetimes

Excess lifetime distributions

Description

Quantile, distribution, density and hazard functions of excess lifetime distribution for threshold exceedances.

Usage

qelife(
  p,
  rate,
  scale,
  shape,
  family = c("exp", "gp", "weibull", "gomp", "gompmake", "extgp", "extweibull", "perks",
    "perksmake", "beard", "beardmake"),
  lower.tail = TRUE
)

pelife(
  q,
  rate,
  scale,
  shape,
  family = c("exp", "gp", "weibull", "gomp", "gompmake", "extgp", "extweibull", "perks",
    "perksmake", "beard", "beardmake"),
  lower.tail = TRUE,
  log.p = FALSE
)

relife(
  n,
  scale = 1,
  rate,
  shape,
  family = c("exp", "gp", "weibull", "gomp", "gompmake", "extgp", "extweibull", "perks",
    "perksmake", "beard", "beardmake")
)

delife(
  x,
  scale = 1,
  rate,
  shape,
  family = c("exp", "gp", "weibull", "gomp", "gompmake", "extgp", "extweibull", "perks",
    "perksmake", "beard", "beardmake"),
  log = FALSE
)

helife(
  x,
  scale = 1,
  rate,
  shape,
  family = c("exp", "gp", "weibull", "gomp", "gompmake", "extgp", "extweibull", "perks",
    "perksmake", "beard", "beardmake"),
  log = FALSE
)

Arguments

p	vector of probabilities
rate	rate parameter(s); for models with Makeham component, the last entry should be part of the rate vector
scale	scale parameter
shape	vector of shape parameter(s).
family	string indicating the parametric model, one of exp, gp, gomp, gompmake, weibull, extgp, extweibull, perks, perksmake, beard and beardmake
lower.tail	logical; if TRUE (default), the lower tail probability \Pr(X \leq x) is returned.
q	vector of quantiles.
n	sample size
log, log.p	logical; if TRUE, values are returned on the logarithmic scale (default to FALSE).

Value

depending on the function type, a vector of probabilities (pelife), quantiles (qelife), density (delife), or hazard (helife). The function relife returns a random sample of size n from the distribution.

longevity documentation built on Nov. 12, 2023, 5:07 p.m.