Has anyone used the LMDI code in STATA? I have been trying to use the LMDI syntax to decompose water decoupling elasticity to its drivers. I am using the following codes (an example): lmdi DecoupIndex = IrrArea Pop GDPReal UrbRate, t(year) over(ProvinceID)

I am receiving this error message: “ERROR: The specified variables can not form an identity”.
Does anyone know what is the problem here? My data is structured as a long panel. For instance, using the setup I can run panel fixed effects regression model.
Thank you very much for reading. Appreciate it if anyone can help.