Dear all,
When I run a simple estimation with this command :
estimation(
_ datafile=data,presample=0,first_obs=1,nobs=97_
_ ,prefilter=1_
_ ,plot_priors=0_
_ ,mh_replic=0000 );_
and when I change the order of varobs, I find a different result ( in vraisemblance and estimation).
It is normal or we need to specify an other option in the estimation program ?
Best
Results should be independent of the ordering of the variables. Could you provide more details so that we can replicate the issue (if there is one)?
Best,
Stéphane.
Yes of course,
I joined an example of mod file and data. When i changed the order of c_obs, y_obs and TI_obs we have different running. It also happens when i have example of model that work correctly.
This problem seems to disappear when i have more observable variable.
Thanks in advance,
Best, data2.m (3.8 KB)
example.mod (8.3 KB)