Hello everyone,
I am running a gsem model using the SEM builder and I keep getting this error: parentheses do not balance. I´ve redone the model several times, adding the paths carefully and keep getting the same message. However, if I run the model from my written code it works perfectly.
This is a simplified example of what I´m doing. When I run the code, it converges quickly. When I build the same paths with the SEM builder, at some points it doesn´t work and the error appears.
gsem (age i.marital_st i.parity i.education preg_BMI i.GWG i.breastfeeding-> i.postpartum_w_change, family(multinomial) link(logit)) ///
(age i.marital_st i.parity i.education preg_BMI i.GWG i.breastfeeding-> bmi_72, family(gaussian) link (identity)) ///
(age i.marital_st i.parity i.education preg_BMI i.GWG i.breastfeeding bmi_72->ln_ldl_72, family(gaussian) link (identity)) ///
(postpartum_w_change-> bmi_72 ln_ldl_72, family(gaussian) link (identity)), nocapslatent
I´m assuming that if the model runs with the written code, it is properly specified. However, I´m not sure if I should worry for not being able to run it using the SEM builder. Is there any reason to prefer using the SEM builder vs writing the code?
Thanks
Diana
Related Posts with Gsem -problems with SEM builder tool
Merging two variables + timeDear Statalist users I have some activity data which asks respondents how much time they spent in p…
Point transparency in spmapHi all, I wonder if its possible control the Point transparency using spmap: Code: spmap relig1 …
Need some helpIm working with a "Encuesta Permanente de Hogares" (EPH) can someone help me with something? …
while loop error, testing overdispersionHello everyone, i have an issue with a loop. I got this from Hilbe's book "modeling count data". I'm…
Looking for a way to do local regressions on data with many zeroesI want to run a nonparametric or semiparametric regression on data which I suspect to be non-linear.…
Subscribe to:
Post Comments (Atom)
0 Response to Gsem -problems with SEM builder tool
Post a Comment