Unexpected interval column can cause error

[racheleclapham]

Hi. I am really enjoying using the package!

I am running into a problem that I can see was fixed in September 2023 (#238/239). I am trying to run pk.nca on my merged concentration and dose data.
ie: my.data.manual<-PKNCAdata(conc_obj,dose_obj,intervals = my.intervals)
my.results.manual<-pk.nca(my.data.manual)

However, I am still receiving the following error:
Caused by error in vapply():
#> ! values must be length 1,
#> but FUN(X[[2]]) result is length 0

I was wondering if you know how to fix this/what I am doing wrong?

Many thanks
Rachel

[billdenney]

Hi Rachel,

I'm glad that you are enjoying PKNCA.

I can't reproduce the issue on my end. Can you please send the full code and data you are using? If the data are proprietary, can you please share the interval definition (my.intervals)?

[racheleclapham]

Hi Bill,

Thank you for replying so quickly!

This is the code I am using:

d_conc<-ptall

d_dose <- d_conc[d_conc$time == 0,]
d_dose$time <- 0

conc_obj <- PKNCAconc(d_conc,
                      +dv~time|group+ID)
dose_obj <- PKNCAdose(d_dose,
                      amt~time|group+ID)

my.intervals<-data.frame(group=c("A","B","C","D","E","F","G","H","I","J"),start=0,end=Inf,cmax=T,ctrough=T, tmax=T, half.life=T,aucinf=T,auclast=T)
my.data.manual<-PKNCAdata(conc_obj,dose_obj,intervals = my.intervals)
my.results.manual<-pk.nca(my.data.manual)
knitr::kable(summary(my.results.manual))

I have also attached the data I am using.

Thank you so much for your help
Rachel

[billdenney]

Hi Rachel,

I have not yet generated a new version for CRAN, so the September fix requires installing from GitHub (remotes::install_github("billdenney/pknca")).

Your issue is identical to the one in #238, the intervals use a column named aucinf rather than aucinf.obs (or similar). This will work with the current version of PKNCA on CRAN:

library(PKNCA)

d_conc<-read.csv("https://github.com/billdenney/pknca/files/13478420/ptall.csv")

d_dose <- d_conc[d_conc$time == 0,]

conc_obj <- PKNCAconc(d_conc,
                      dv~time|group+ID)
dose_obj <- PKNCAdose(d_dose,
                      amt~time|group+ID)

my.intervals<-data.frame(start=0,end=Inf,cmax=T,ctrough=T, tmax=T, half.life=T,aucinf.obs=T,auclast=T)
my.data.manual<-PKNCAdata(conc_obj,dose_obj,intervals = my.intervals)
my.results.manual<-pk.nca(my.data.manual)
knitr::kable(summary(my.results.manual))

A couple of notes to help with future analyses:

• There was an extra "+" before "dv" in the PKNCAconc formula. It doesn't hurt anything, but it isn't required. (I made that change in the above code.)
• If all of your groups need the same NCA parameters calculated, then you do not need the group definitions in the intervals. (I made that change in the above code.)

[racheleclapham]

Thank you so much, I really appreciate your help!