relimp (version 0.5)

vcov: Covariance Matrix of Model Estimates

Description

vcov is a generic function from the MASS package, for obtaining the estimated variance-covariance matrix of parameter estimates in a fitted model. Methods are available at least for models of class lm, glm, nls, polr, multinom, coxph, survreg, gls, lme.

Usage

## S3 method for class 'coxph':
vcov(object, ...)

Arguments

object

A fitted model object

...

Additional arguments for method functions (e.g., for models of class glm, the dispersion argument may sometimes be used.

Value

• A matrix.

Details

Extracts the estimated variance-covariance matrix of the estimator which produced the estimated coefficients in coef(object).

See Also

vcov

Examples

library(MASS)
library(survival)
data(ovarian)
modl <- survreg(Surv(futime, fustat) ~ ecog.ps + rx,
                ovarian, dist='weibull',scale=1)
vcov(modl)