VPC for censored data

[jfstanding]

I am having trouble creating a VPC for the continuous part of a censored fit. Here is an example:

library(nlmixr2)
library(ggplot2)
# create theo censored
theo_cens <- theo_sd
theo_cens$cens <- 0
theo_cens$cens[theo_cens$DV <= 1] <- 1
theo_cens$DV[theo_cens$DV <= 1 & theo_cens$AMT == 0] <- 1

m1 <- function() {
  ini({
    tka <- 0.5
    tcl <- -3.2
    tv <- -1
    eta.ka ~ 1
    eta.cl ~ 2
    eta.v ~ 1
    add.err <- 0.1
  })
  model({
    ka <- exp(tka + eta.ka)
    cl <- exp(tcl + eta.cl)
    v <- exp(tv + eta.v)
    linCmt() ~ add(add.err)
  })
}
fit1 <- nlmixr(m1, theo_cens, 
               est = "focei", table = tableControl(npde = TRUE, censMethod = "cdf"))
vpcPlot(fit = fit1)
# fails
#> Error in `dplyr::summarise()`:
#>! Problem while computing `obs5 = quantile_cens(dv, pi[1], limit = limit, cens = cens)`.
#>ℹ The error occurred in group 1: strat = 1, bin = 1.
#> Caused by error in `if (cens %in% c("left", "lower", "bloq", "loq", "lloq")) ...`:
#> ! the condition has length > 1
#> Run `rlang::last_error()` to see where the error occurred.
# 
# Note:
# vpcPlot(fit = fit1, lloq = 1)
# fit1$dataLloq
# Correctly identifies lloq
#
vpcPlot(fit = fit1, cens = TRUE)
# works for the censored part

The censored VPC cens = TRUE works but the continuous part does not.

[mattfidler]

Thanks for reporting this @jfstanding

This bug has been fixed. You can probably see the fix by installing nlmixr2plot as follows:

devtools::install_github("nlmixr2/nlmixr2plot")