premature temination
157 views
Skip to first unread message
aaditya bhamidipati
Jun 21, 2020, 8:16:25 AM6/21/20
to Biogeme
Dear Prof Bierlaire and forum members,
i have been trying to perform a simultaneous estimation on an ICLV model with three latent variables, having 3 indicators for one, two for the second and 1 for the third. though the relative gradient criteria has been met i.e. relative gradient = 4.9e-304 <= 6.1e-06, the results dont seem to be appropriate as the p value is exactly zero for all the parameters , the robust p being one and correlation of coefficients is 1 for all the parameters. Can anyone please guide me in resolving the error? I have the script attached for reference.
i have been trying to perform a simultaneous estimation on an ICLV model with three latent variables, having 3 indicators for one, two for the second and 1 for the third. though the relative gradient criteria has been met i.e. relative gradient = 4.9e-304 <= 6.1e-06, the results dont seem to be appropriate as the p value is exactly zero for all the parameters , the robust p being one and correlation of coefficients is 1 for all the parameters. Can anyone please guide me in resolving the error? I have the script attached for reference.
Thanks
Aaditya
Bierlaire Michel
Jun 22, 2020, 2:51:37 AM6/22/20
to aadit...@gmail.com, Bierlaire Michel, Biogeme
Why did you fix the sigma’s to 1?
--
You received this message because you are subscribed to the Google Groups "Biogeme" group.
To unsubscribe from this group and stop receiving emails from it, send an email to biogeme+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/biogeme/7a120f32-6df5-456c-b1fc-80224ff0fb88o%40googlegroups.com.
<ICLV full(1).py>
Reply all
Reply to author
Forward
0 new messages