Error running using rpy2 for survival analysis

Question

I'm trying to perform survival analysis with R with rpy2 in ipython.

The following formulation fails:

%Rpush df
surv_fit = surv.coxph("Surv(start, end, status) ~ x", data=df)
> RRuntimeError: Error: $ operator is invalid for atomic vectors

But if I do it all using r magic, it works:

%Rpush df
%R surv.fit <- coxph(Surv(start, end, status) ~ x, data=df)

Is there an issue with the Surv object?

Answer 1

This is because %Rpush is pushing the variable to R.

Your first example would then have to be: surv_fit = rpy2.robjects.r(""" surv.coxph(Surv(start, end, status) ~ x, data=df) """)