Re: data
Hi Andre,
It seems like you got the help you needed. I just want to bring to your attention that the R package NMdata provides a function called NMcheckData that can scan a data set for compatibility with Nonmem and additional pitfalls for PK or PK/PD data sets. Admittedly, it would report a finding if MDV=1 and DV<0.
Sometimes the problems occur because the the $INPUT section in the control stream has a different order of columns than the data set. If you pass a control stream as
NMcheckData(file=<path-to-control-stream>)
it will read the data as defined in the control stream, compare to column names (if found) in the input data file, and then turn to checking the data set for nonmem compatibility. So if something weird is going on, that can very often point you to where and why.
Best,
Philip
Quoted reply history
On 9/5/22 10:41, [email protected] wrote:
> All:
>
> I have data which looks like this with the line number inserted in ( ).
>
> WSUBJ TIME CP MDV CTIME DAY AMT EVID DV
>
> (1)1 0 0 0 0 1 0.001 1 0.153846
>
> (12)2 0 0 0 0 1 0.001 1 -0.30769
>
> (21)3 0 0 0 0 1 0.001 1 1.076923
>
> I get the following warning with each run.
>
> (DATA WARNING 5) RECORD 1, DATA ITEM 9, CONTENTS 0.1538
>
> THE DV DATA ITEM IS POSITIVE, BUT THE MDV ITEM IS 1
>
> (DATA WARNING 5) RECORD 21, DATA ITEM 9, CONTENTS 1.0769
>
> THE DV DATA ITEM IS POSITIVE, BUT THE MDV ITEM IS 1
>
> Each is a dosing for PD; however, it does not happen for record 12, also a dosing record. Can someone tell me why, and how to address this issue?