mlreg: ML proportional hazards regression

In eha: Event History Analysis

ML proportional hazards regression

Description

Maximum Likelihood estimation of proportional hazards models. Is deprecated, use coxreg instead.

Usage

mlreg(
  formula = formula(data),
  data = parent.frame(),
  na.action = getOption("na.action"),
  init = NULL,
  method = c("ML", "MPPL"),
  control = list(eps = 1e-08, maxiter = 10, n.points = 12, trace = FALSE),
  singular.ok = TRUE,
  model = FALSE,
  center = TRUE,
  x = FALSE,
  y = TRUE,
  boot = FALSE,
  geometric = FALSE,
  rs = NULL,
  frailty = NULL,
  max.survs = NULL
)

Arguments

formula	a formula object, with the response on the left of a ~ operator, and the terms on the right. The response must be a survival object as returned by the Surv function.
data	a data.frame in which to interpret the variables named in the formula.
na.action	a missing-data filter function, applied to the model.frame, after any subset argument has been used. Default is options()$na.action.
init	vector of initial values of the iteration. Default initial value is zero for all variables.
method	Method of treating ties, "ML", the default, means pure maximum likelihood, i.e, data are treated as discrete. The choice "MPPL" implies that risk sets with no tied events are treated as in ordinary Cox regression. This is a cameleont that adapts to data, part discrete and part continuous.
control	a list with components eps (convergence criterion), maxiter (maximum number of iterations), and silent (logical, controlling amount of output). You can change any component without mention the other(s).
singular.ok	Not used.
model	Not used.
center	Should covariates be centered? Default is TRUE
x	Return the design matrix in the model object?
y	return the response in the model object?
boot	No. of bootstrap replicates. Defaults to FALSE, i.e., no bootstrapping.
geometric	If TRUE, the intensity is assumed constant within strata.
rs	Risk set? If present, speeds up calculations considerably.
frailty	A grouping variable for frailty analysis. Full name is needed.
max.survs	Sampling of risk sets?

Details

Method ML performs a true discrete analysis, i.e., one parameter per observed event time. Method MPPL is a compromize between the discrete and continuous time approaches; one parameter per observed event time with multiple events. With no ties in data, an ordinary Cox regression (as with coxreg) is performed.

Value

A list of class c("mlreg", "coxreg", "coxph") with components

coefficients	Fitted parameter estimates.
var	Covariance matrix of the estimates.
loglik	Vector of length two; first component is the value at the initial parameter values, the second componet is the maximized value.
score	The score test statistic (at the initial value).
linear.predictors	The estimated linear predictors.
residuals	The martingale residuals.
hazard	The estimated baseline hazard.
means	Means of the columns of the design matrix.
w.means	Weighted (against exposure time) means of covariates; weighted relative frequencies of levels of factors.
n	Number of spells in indata (possibly after removal of cases with NA's).
events	Number of events in data.
terms	Used by extractor functions.
assign	Used by extractor functions.
wald.test	The Walt test statistic (at the initial value).
y	The Surv vector.
isF	Logical vector indicating the covariates that are factors.
covars	The covariates.
ttr	Total Time at Risk.
levels	List of levels of factors.
formula	The calling formula.
call	The call.
bootstrap	The bootstrap sample, if requested on input.
sigma	Present if a frailty model is fitted. Equals the estimated frailty standard deviation.
sigma.sd	The standard error of the estimated frailty standard deviation.
method	The method.
convergence	Did the optimization converge?
fail	Did the optimization fail? (Is NULL if not).

Warning

The use of rs is dangerous, see note above. It can however speed up computing time.

Note

This function starts by creating risksets, if no riskset is supplied via rs, with the aid of risksets. This latter mechanism fails if there are any NA's in the data! Note also that it depends on stratification, so rs contains information about stratification. Giving another strata variable in the formula is an error. The same is ok, for instance to supply stratum interactions.

Note futher that mlreg is deprecated. coxreg should be used instead.

Author(s)

Göran Broström

References

Broström, G. (2002). Cox regression; Ties without tears. Communications in Statistics: Theory and Methods 31, 285–297.

See Also

coxreg, risksets

Examples

 dat <- data.frame(time=  c(4, 3,1,1,2,2,3),
                status=c(1,1,1,0,1,1,0),
                x=     c(0, 2,1,1,1,0,0),
                sex=   c(0, 0,0,0,1,1,1))
 mlreg( Surv(time, status) ~ x + strata(sex), data = dat) #stratified model
 # Same as:
 rs <- risksets(Surv(dat$time, dat$status), strata = dat$sex)
 mlreg( Surv(time, status) ~ x, data = dat, rs = rs) #stratified model
 

eha documentation built on Oct. 7, 2024, 1:17 a.m.