Using a very long variable names in the Surv() portion of the formula crash the test for "equal" outcomes in the anova function (the reason is the outcome is split across multiple lines when parsed from the formula.
library(survival)
ovarianB <- ovarian
ovarianB$futime________________________long <- ovarianB$futime
fit <- coxph(Surv(time = futime________________________long, event = fustat, type = "right") ~ resid.ds *rx + ecog.ps, data = ovarianB)
fit2 <- coxph(Surv(time = futime________________________long, event = fustat, type = "right") ~ resid.ds +rx + ecog.ps, data = ovarianB)
anova(fit2,fit)
Using a very long variable names in the
Surv()
portion of the formula crash the test for "equal" outcomes in theanova
function (the reason is the outcome is split across multiple lines when parsed from the formula.