[R] Calculation of the median in survfit()

[Christoph Scherber]

Dear R users,

Can anyone tell me how the medians in survfit() are computed? I´ve 
looked it up in the source code (print.survfit.s version 4.19 07/09/00), 
but I´m not a programmer...;-)

Especially, I´d like to know what the pfun() function inside 
print.survfit.s() works.

The help file does describe the calculation of the median, but I´d like 
to know if and how this differs from just directly calculating the 
median of the survival times (e.g. using tapply(time,factor,median)).

I would be most grateful for any help. Thanks a lot!
Christoph

###
[I am using R 2.1.1 on WinXP with the latest version of the survival 
package]