Feature request: `survFit()` with weights

[hfrick]

Could we get survFit() to accept a weighted fit? The "yet" in the error message suggests you might be open to this? 🙏

Context: we would like to support case weights for mboost in the tidymodels framework but if only model fitting but not prediction (what we use survFit() for), we get surprised users like in tidymodels/censored#363. I think that's a fair assumption from that user so I've disabled case weights for mboost for the moment but I'd love to bring it back!

library(mboost)
#> Warning: package 'mboost' was built under R version 4.6.1
#> Loading required package: parallel
#> Loading required package: stabs
library(survival)
#> Warning: package 'survival' was built under R version 4.6.1

lung2 <- na.omit(lung[, c("time", "status", "age", "ph.ecog")])
set.seed(1)
wts <- runif(nrow(lung2))

# A weighted boosted Cox model fits without complaint:
fit <- blackboost(
  Surv(time, status) ~ age + ph.ecog,
  data = lung2,
  family = CoxPH(),
  weights = wts
)

# ... but survFit() rejects it:
survFit(fit, newdata = lung2[1:3, ])
#> Error in `survFit.mboost()`:
#> ! survFit cannot (yet) deal with weights

^{Created on 2026-07-23 with reprex v2.1.1}