Hi Pascal,
This looks like a bug (in Nonmem or in your code) to me. With TOL=16,
there should be no numerical problems with ODE. Could you provide more
details (code with the initial conditions + sample of the data for one
subject where you have this problem)?
Thanks
Leonid
-
Hi Leonid,
Thanks for the additional suggestion to use ADVAN13. I was able to
increase TOL up to 16, SIGL to 14, but still have the same biases for the
moderate to almost flat initial slope after baseline when using dummy
points spaced every 1 unit of time. When I reduce number of dummy points