Use of weights with cox() family

[cccnrc]

I already started a topic on stackoverflow: Using IPTW weights inside a Bayesian brm() cox model

I can create a Bayesian coxph model with brms:

library(brms)

### use brms package dataframe
data("kidney")

### create random propensity scores to simulate the database
kidney$sw <- runif( nrow(kidney), min=0.5, max=2.5 )

### create the brm model
model <- brms::brm(
  time | cens(censored) ~ sex,
  data = kidney,
  family = cox()
)

and everything works fine. I tried to implement weights but I cannot figure out the syntax:

model <- brms::brm(
  Surv( time, censored ) | weights(sw) ~ sex,
  data = kidney,
  family = cox()
)
Error in Ops.Surv(y, 0) : Invalid operation on a survival time

How can I specify weights if I am using a cox() syntax?

[paul-buerkner]

The following works:

model <- brms::brm(
time | cens(censored) + weights(sw) ~ sex,
data = kidney,
family = cox()
)