Calculate observed survival informations

This functions calculate some survival information (e.g. survival rate, mortality rate).

observed_survival(x, f, times)

observed_mortality(x, f, times)

observed_events(x, f, times, cumulative = FALSE)

Arguments

x

Surv, object

f

factor, strata/splitting factor

times

numeric, information at times

cumulative

logical, if TRUE (default FALSE) and multiple times are given the cumulative events are reported instead of the events in the time periods.

Value

double, with an element for each level in f, if times is a numeric of length 1; or a matrix with as much rows as elements in times and a column for each level in f.

Details

In contrast to survival::summary.survfit() these functions do not remove empty strata (e.g. important for bootstrapping).

Examples

sv <- Surv(1:4, c(0, 1, 1, 0))
f <- factor(c("a", "c", "a", "c"), levels = c("a", "b", "c"))
observed_survival(sv, times = 3)
#> [1] 0.3333333
observed_survival(sv, f = f, times = 3)
#>   a   b   c 
#> 0.0  NA 0.5 
observed_survival(sv, f = f, times = 1:4)
#>   a  b   c
#> 1 1 NA 1.0
#> 2 1 NA 0.5
#> 3 0 NA 0.5
#> 4 0 NA 0.5

observed_mortality(sv, f = f, times = 3)
#>   a   b   c 
#> 1.0  NA 0.5 

observed_events(sv, f = f, times = 3)
#>  a  b  c 
#>  1 NA  1 
observed_events(sv, f = f, times = 1:4)
#>   a  b c
#> 1 0 NA 0
#> 2 0 NA 1
#> 3 1 NA 0
#> 4 0 NA 0
observed_events(sv, f = f, times = 1:4, cumulative = TRUE)
#>   a  b c
#> 1 0 NA 0
#> 2 0 NA 1
#> 3 1 NA 1
#> 4 1 NA 1