I am doing a panel data analysis in R Software which is truncated with more than 700 firms and a 20 year sample period. While using the "trunc" option in the pldv function, it gives this error:
modelo
not clear what software you're using but it would seem the hessian is the optimisation function of a LL? in which case the censored regression puts the mass points of the truncated obs at the point, so its sounds then less surprising that the matrix is nolonger exactly singular no? maybe contemplating this will help solve it. good luck. :)
I am not conversant with this software, but with the mention of the term "singular", I suggest u check if two of your variables are highly correlated. Drop one of the suspected variables and re-do your regression, then, check the outcome.
- Badges
- Science topic