Compute Weibull model
For a Weibull model with covariates, we can compute the survival curve just as we did for the Weibull model without covariates. The only thing we need to do is specify the covariate values for a given survival curve in the predict()
function. This can be done with the argument newdata
.
This exercise is part of the course
Survival Analysis in R
Exercise instructions
- Compute a Weibull model for the
GBSG2
dataset with covariatehorTh
to analyze the effect of hormonal therapy on the survival of patients. - Compute the survival curve for patients who receive hormonal therapy.
- Take a look at the survival curve with
str()
.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Weibull model
wbmod <- survreg(Surv(___, ___) ~ ___, data = GBSG2)
coef(wbmod)
# Retrieve survival curve from model
surv <- seq(.99, .01, by = -.01)
t_yes <- predict(___, type = "quantile", p = ___,
newdata = data.frame(horTh = ___))
# Take a look at survival curve
str(t_yes)