Fast Cox PH regression — phreg • mets

Fast Cox PH regression Robust variance is default variance with the summary.

phreg(formula, data, offset = NULL, weights = NULL, ...)

Arguments

formula: formula with 'Surv' outcome (see coxph)
data: data frame
offset: offsets for Cox model
weights: weights for Cox score equations
...: Additional arguments to lower level funtions

Details

influence functions (iid) will follow numerical order of given cluster variable so ordering after $id will give iid in order of data-set.

Author

Klaus K. Holst, Thomas Scheike

Examples

data(TRACE)
dcut(TRACE) <- ~.
out1 <- phreg(Surv(time,status==9)~vf+chf+strata(wmicat.4),data=TRACE)
## robust standard errors default 
summary(out1)
#> 
#>     n events
#>  1878    958
#> coeffients:
#>     Estimate     S.E.  dU^-1/2 P-value
#> vf  0.452306 0.136473 0.111038   9e-04
#> chf 0.931822 0.074226 0.074650   0e+00
#> 
#> exp(coeffients):
#>     Estimate   2.5%  97.5%
#> vf    1.5719 1.2030 2.0540
#> chf   2.5391 2.1954 2.9367
#> 
#> 

par(mfrow=c(1,2))
plot(out1)

## computing robust variance for baseline
rob1 <- robust.phreg(out1)
plot(rob1,se=TRUE,robust=TRUE)


## taking iid decomposition of regression parameters
betaiiid <- lava::iid(out1)

## making iid decomposition of baseline at a specific time-point
Aiiid <- mets::IIDbaseline.phreg(out1,time=30)