Cross Validated is a question and answer site for people interested in statistics, machine learning, data analysis, data mining, and data visualization. It only takes a minute to sign up.
Sign up to join this community
I have the following Cox PH model:
(Time, Event) ~ X + Y + Z
I would like to get the predicted hazard rates (i am talking about hazard rates NOT hazard ratios) given specific values of X, Y, Z. I know the muhaz R package can calculate the observed hazard rates, but I am interested in the predicted model.
Is there a way to do this in R?
The function in the R survival package to get the baseline hazard rate is basehaz.
Then you need to multiply it for the various $e^{\beta}$ to get the specific hazard rate given the coefficients you have found.
A simple example may help:
library(survival) #survival analysis
library(eha) #used for data
data(oldmort) #create the data
# Create surv data set
mort <- Surv(time=oldmort$enter,time2=oldmort$exit,event=oldmort$event)
reg_fit <- coxph(formula=mort~oldmort$sex)
summary(reg_fit)
# Now get baseline curve
baseline <- basehaz(reg_fit)
# Draw baseline hazard (that's male)
plot(baseline$time, baseline$hazard, type='l',main="Hazard rates")
# Draw female hazard
lines(baseline$time, exp(-0.1929)*baseline$hazard, col="blue")
the $\exp(-0.1929)=0.8245$ is copied and pasted from the summary results of cox regression, summary(reg_fit), implying that the females in the data have hazard rates $18\%$ lower than the males.
The function basehaz (from the previous answer) provides the cumulative hazard, not the hazard function (the rate). I believe that question was about the hazard function. Estimating the hazard function would require specification of the type of smoothing (like in density estimation). The Muhaz R package can do this for one sample data. I am not aware of a function that can do this for the baseline hazard in the Cox model. I also need this. I think I will need to smooth the cumulative hazard on my own.
