topic Re: problem with NLMIXED Execution error for observation 1 in Statistical Procedures

problem with NLMIXED Execution error for observation 1

alfa95 — Sun, 26 Jan 2020 17:17:49 GMT

Hello everybody,

I am doing non linear mixed model.

my code is as follow:

proc nlmixed data=dataq1prev2 qpoints=4;
parms  eta=2.02 etadiff=-0.32 tau=6.24
taudiff=0.9 gamma=-0.002 gdiff=0 sigma2=0.0007 
d11=0.0022 d22=0.0009 d33=0.14;
num = ((eta + etadiff * group + n));
den = ((tau + t+ taudiff * group );
mean = num/den + gamma + gdiff * group ;
model reponse~ normal(mean,sigma2);
random t n ~ normal([0, 0],[d11,0,d22])
subject=vog out=eb;
run;

I get error saying "NOTE: Execution error for observation 1."

I will very appreciate any help, thank you

Re: problem with NLMIXED Execution error for observation 1

PaigeMiller — Sun, 19 Jan 2020 22:39:21 GMT

Please describe the problem.

Re: problem with NLMIXED Execution error for observation 1

alfa95 — Sun, 19 Jan 2020 22:57:54 GMT

@PaigeMiller I edited the post, thank you

Re: problem with NLMIXED Execution error for observation 1

PaigeMiller — Sun, 19 Jan 2020 23:12:50 GMT

So step #1

You look at the data yourself, and make sure there are no missing values in observation 1, and that all mathematical operations that you are trying to do on this data are defined (for example, if you have a negative number you can't raise it to a non-integer power).

Re: problem with NLMIXED Execution error for observation 1

alfa95 — Sun, 19 Jan 2020 23:26:04 GMT

@PaigeMiller

thank you