| coxphw.control {coxphw} | R Documentation |
Ancillary arguments for controlling coxphw fits
Description
This is used to set various numeric parameters controlling a Cox model fit using coxphw.
Typically it would only be used in a call to coxphw.
Usage
coxphw.control(iter.max = 200, maxhs = 5, xconv = 1e-4, gconv = 1e-4, maxstep = 1,
round.times.to = 0.00001, add.constant = 0, pc = TRUE, pc.time = TRUE,
normalize = TRUE)
Arguments
iter.max |
maximum number of iterations to attempt for convergence. Default is 200. |
maxhs |
maximum number of step-halvenings per iterations. Default is 5. The increments of the
parameter vector in one Newton-Rhaphson iteration step are halved, unless the new
pseudo-likelihood is greater than the old one, maximally doing |
xconv |
specifies the maximum allowed change in standardized parameter estimates to declare convergence. Default is 0.0001. |
gconv |
specifies the maximum allowed change in score function to declare convergence. Default is 0.0001. |
maxstep |
specifies the maximum change of (standardized) parameter values allowed in one iteration. Default is 1. |
round.times.to |
rounds survival times to the nearest number that is a multiple of
|
add.constant |
this number will be added to all times. It can be used if some survival times are exactly 0. Default is 0. |
pc |
if set to TRUE, it will orthogonalize the model matrix to speed up convergence. Default is TRUE. |
pc.time |
if set to TRUE, it will orthogonalize time-dependent covariates (i.e., interactions of covariates with functions of time) to speed up convergence. Default is TRUE. |
normalize |
if set to TRUE, weights are normalized such that their sum is equal to the number of events. This may speed up or enable convergence, but has no consequences on the estimated regression coefficients. |
Value
A list containing the values of each of the above constants
Author(s)
Daniela Dunkler