Quantiles for Parametric Survival/Censored Distributions

DESCRIPTION:

Computes quantiles for parametric survival distributions.

USAGE:

qftdist(p=0.1, location=0, scale=1, distribution="extreme", link="log",  
fit=NULL, X=NULL, coverage=0.9, data=NULL, se.fit=T, conf.interval=T) 

OPTIONAL ARGUMENTS:

p
vector of probabilities for which quantiles are desired. Argument p cannot be of length zero.
location
numeric value specifying location parameter for the quantiles.
scale
numeric value specifying scale parameter for the quantiles.
distribution
character string specifying the distribution. This may be one of "weibull", "extreme", "lognormal", "normal", "loggaussian", "gaussian", "loglogistic", "logistic", "logexponential", "exponential", "lograyleigh", and "rayleigh".
fit
a fitted "censorReg" object. This may be either an object that inherits from class "censorReg", or a list of objects that inherit from class "censorReg". If present, arguments location, scale, and distribution are ignored.
X
data.frame containing the covariate values at which quantiles are to be computed. Not used if argument fit is NULL. The values of covariates for all stratification, offset, and independent variables must be specified.
coverage
numeric value giving the coverage desired for the confidence intervals.
data
used when calling qftdist from within another routine. In this case the data argument must be specified.
se.fit
logical value indicating whether standard errors of the estimate are to be returned. Default is T.
conf.interval
logical value indicating whether confidence intervals for the estimates are to be returned. If T, upper and lower boundaries of confidence intervals are returned.

VALUE:

A matrix of quantiles and confidence intervals. If X is input, then a list of matrices of quantiles and confidence intervals is computed, one for each element in the data frame X.

DETAILS:

When fit is input, threshold is added to the estimated quantile. Offsets, if any, are also incorporated into the quantile estimate.

SEE ALSO:

EXAMPLES:

qftdist(p=c(0.01, 0.05, 0.1, 0.5, 0.9, 0.95, 0.99), distribution="normal") 
a <- censorReg(censor(days,event)~voltage,data=capacitor) 
qftdist(p=c(0.01, 0.05, 0.1, 0.5), fit=a, X=20:32)