eha (version 2.4-5)

phreg.fit: Parametric proportional hazards regression

Description

This function is called by phreg, but it can also be directly called by a user.

Usage

phreg.fit(X, Y, dist, strata, offset, init, shape, control, center = NULL)

Arguments

X

The design (covariate) matrix.

Y

A survival object, the response.

dist

Which baseline distribution?

strata

A stratum variable.

offset

Offset.

init

Initial regression parameter values.

shape

If positive, a fixed value of the shape parameter in the distribution. Otherwise, the shape is estimated.

control

Controls convergence and output.

center

Deprecated (not used). Kept for backward copability. Results are reported as is, no centering.

Value

coefficients

Estimated regression coefficients plus estimated scale and shape coefficients, sorted by strata, if present.

var

Variance-covariance matrix

loglik

Vector of length 2. The first component is the maximized loglihood with only scale and shape in the model, the second the final maximum.

score

Score test statistic at initial values

linear.predictors

Linear predictors for each interval.

means

Means of the covariates

conver

TRUE if convergence

fail

TRUE if failure

iter

Number of Newton-Raphson iterates.

n.strata

The number of strata in the data.

Details

See phreg for more detail.

See Also

phreg