This function compares survival curves as modeled with do_kpm().
It outputs a contingency table and a Chisq measure of difference.
Arguments
- kds
data set of a survival model such as
do_kpm()- cond
character of experimental condition variable in the data
- test_type
numeric (0 or 1) parameter that controls the type of test (0 means rho = 0; log-rank, 1 means rho = 1; Peto & Peto Wilcox)
Value
Returns survival test results as called from survival::survdiff().
Examples
kpm_est <- do_kpm(add_dropout_idx(dropRdemo, 3:54))
get_survdiff(kpm_est$d, "experimental_condition", 0)
#> Call:
#> survdiff(formula = f, data = kds, rho = test_type)
#>
#> N Observed Expected (O-E)^2/E (O-E)^2/V
#> experimental_condition=11 72 72 72.1 0.000148 0.000358
#> experimental_condition=12 57 56 50.1 0.689840 1.587210
#> experimental_condition=21 56 56 58.2 0.084353 0.185905
#> experimental_condition=22 61 61 64.6 0.196369 0.443239
#>
#> Chisq= 1.8 on 3 degrees of freedom, p= 0.6
get_survdiff(kpm_est$d, "experimental_condition", 1)
#> Call:
#> survdiff(formula = f, data = kds, rho = test_type)
#>
#> N Observed Expected (O-E)^2/E (O-E)^2/V
#> experimental_condition=11 72 50.5 49.8 0.00726 0.0278
#> experimental_condition=12 57 44.9 38.3 1.12936 3.9254
#> experimental_condition=21 56 37.5 39.3 0.08966 0.3173
#> experimental_condition=22 61 37.9 43.2 0.64983 2.3712
#>
#> Chisq= 5.1 on 3 degrees of freedom, p= 0.2