Lower limit of the pointwise confidence interval of S(t) (where S(t) = 1, 1 for each confType; with "log-log"
as survival 3.8-12, NA before it)
Per stratum, [median, lower limit, upper limit] (NaN for R's NA)
The strata of median
Number censored at the time (R's n.censor)
Number of events at the time (R's n.event)
Number at risk just before the time (R's n.risk)
Standard error of S(t) (S times the standard error of log S)
Alias of stdErr: the same array.
Stratum of each row ("group=<label>", or "all" for a single curve)
Kaplan-Meier estimate S(t)
Time of each row (the distinct times of the stratum, censored ones included)
Upper limit of the pointwise confidence interval of S(t)
Result of survfit: the rows of R's
summary(fit, censored = TRUE)and the medians.