harrelfe / rms

Regression Modeling Strategies
https://hbiostat.org/R/rms
Other
172 stars 48 forks source link

Accidentally turning on backward step-down procedure in `validate` and `calibrate` #142

Open mzobeck opened 5 months ago

mzobeck commented 5 months ago

I was trying to set the number of repetitions represented by argument B in the validate and calibrate functions, but I mistakenly specified b instead (e.g. b = 300). I found this mistake activates the same behavior as bw = T, which I don't think is intended.

Example below:

library(rms)
library(survival)

mod <- cph(Surv(time, status) ~ id + rx + sex + age + obstruct + perfor + adhere + nodes + differ + extent + surg + node4,data = colon, surv = T, x = T, y = T)

validate(mod, B = 100)
validate(mod, b = 100)

calibrate(mod, B = 100, u = 360)
calibrate(mod, b = 100, u = 360)