How can we pull in ETAs and also evaluate ERR() and/or EPS() in $ERROR?
Here is what NMsim_known (the function or "method" NMsim provides to simulate subjects with known ETAs) does and why. It pulls in ETAs from the phi file using an $ETAS statement like this:
How can we pull in ETAs and also evaluate
ERR()
and/orEPS()
in$ERROR
?Here is what NMsim_known (the function or "method" NMsim provides to simulate subjects with known ETAs) does and why. It pulls in ETAs from the phi file using an $ETAS statement like this:
As far as I understand that has to be accompanied by a
$ESTIMATION
step that specifiesFNLETA=2
like this:I have not been able to get
$SIM ONLYSIM
to work with this so instead it just does:This however means that in the following from $ERROR we don't get the
ERR(1)
andERR(2)
simulated and soY=IPRED.
Meaning currently, NMsim_known does not provide simulation with residual variability.