Compute Residuals for a Parametric Censored Regression Model

DESCRIPTION:

Computes four types of residuals from a censored regression model object of class censorReg.

USAGE:

residuals.censorReg(object, type="standardized") 

This is a method for the function residuals for objects inheriting from class "censorReg" or "censorRegList". However, as several types of residuals are available for these classes, there is an additional optional argument type.

REQUIRED ARGUMENTS:

object
an object inheriting from class "censorReg" or "censorRegList", representing a parametric survival or censored regression model. Typically this is the output from the censorReg function.

OPTIONAL ARGUMENTS:

type
type of residuals, with choices "standardized", "deviance", "pearson", or "matrix".

VALUE:

a vector of censored residuals is returned.

The standardized residuals are the standardized Cox-Snell censored residuals described in Meeker and Escobar (1998). These are identical to the Pearson residuals for non-logged distributions except for censored values. The sum of squared deviance residuals add up to the deviance. The Pearson residuals are standardized residuals on the scale of the response.

REFERENCES:

Meeker, W. M. and Escobar, L. A. (1998). Statistical Methods for Reliability Data. New York: Wiley.

EXAMPLES:

fit <- censorReg(censor(days, event) ~ voltage, data=capacitor2, 
    weights=weights) 
rr <- residuals(fit, type="standardized")