Closed DavidJesse21 closed 4 months ago
The survfit and residuals.survfit functions are in the midst of a major rewrite. (An issue with case weights, which was unfortunately in the interior of the underlying C code, i.e., the hardest to fix). One side effect is that the mapply is no more. Hopefully fully debugged and posted soon.
Number 237 has been repaired in the most recent update
Hi,
so I found a hack to circumvent #237 at the moment. My motivation for using
survival::pseudo()
was to use them to obtain bootstrap estimates of the covariance matrix of the resulting regression model. While programming a function that would do that, I encountered another issue, which in this case is more precisely located atrsurvpart1()
. I will first provide a code snippet, which illustrates/reproduces this issue:So I assume this piece of code causes the error:
I don't know what the expected output of this should usually look like (in terms of dimensions etc.) but I assume inside of
mapply()
we should setSIMPLIFY = FALSE
such that the output will be a list in any case. This might also be the case for other lines/pieces of code that use this function.Just as a note: I don't know in which particular cases these errors happen and when they don't. Initially I did not find anything suspicious about the bootstrap sample that caused this error.
Best regards David