coxaalenic.control: Control a Cox-Aalen model fit

Description

Set parameters controlling the model fit returned by coxaalenic.

Usage

coxaalenic.control(eps = 1e-07, eps.norm = c("max", "grad"),
                   iter.max = 5000, armijo = 1/3, coef.typ = 1,
                   coef.max = 10, trace = FALSE, thread.max = 1)

Arguments

eps

threshold value for the norm used to measure convergence in the parameter estimates.

eps.norm

a character string identifying the norm to use in the convergence criteria---either the maximum norm between the current and previous parameter values (eps.norm = "max") or the inner product between the current value and the score

iter.max

maximum number of iterations to attempt. This ensures that coxaalenic will eventually exit, even when the stopping rule is not met. A warning is issued whenever the estimation routine has s

armijo

a scale factor in (0, 1/2) for Armijo's (1966) rule---a step-halving line search used to ensure that each iteration achieves an adequate increase in the log-likelihood. The model fit is not particularly sensitive to this value.

coef.typ

a scalar or vector of typical (absolute) values for the multiplicative regression coefficient.

coef.max

a scalar or vector of probable upper bounds for the multiplicative regression coefficient. This and the coef.typ arguments tune variance estimation via the curvature in the profile log-likelihood.

trace

a logical value indicating that CPLEX should print its results to the screen.

thread.max

maximum number of CPU threads to allocate to CPLEX. Default value disables multithreading. A value of zero allows CPLEX to set the number of threads automatically. Any value exceeding the total number of logical cores on the system is reset

Value

A list of the above arguments with their final values.

References

Boruvka, A. and Cook, R. J. (2014a) A Cox-Aalen model for interval-censored data.

Armijo, L. (1966) http://dx.doi.org/10.2140/pjm.1966.16.1{ Minimization of functions having Lipschitz continuous first partial derivatives.} Pacific Journal of Mathematics 16, 1--3.

Examples

Run this code

if (is.loaded("coxaalenic", "coxinterval"))
coxaalenic(Surv(left, right, type = "interval2") ~ prop(treat),
           data = cosmesis, control = coxaalenic.control(iter.max = 2,
           trace = TRUE))