I have taken first partial derivatives from the same reliable source, where the model was taken. That is, mathematically they must be correct. But it does not matter, those derivatives do not work. Related to this issue, I have some questions, that some expert can hopefully answer. I know that taking derivatives w.r.t. explanatory variables b, Stata silently multiplies the corresponding expression with the variable x, so I do not have to do it. I’m wondering if there are some other silent moves too, which I need to know, to get these derivatives work correctly?
In case of panel data, partial derivatives from explanatory variables b are observation specific, but from variance parameter (which is a scalar), they are panel specific. If the model is heteroscedastic and variance includes explanatory variables, is it still panel specific?
In case of panel data, partial derivatives from explanatory variables b are observation specific, but from variance parameter (which is a scalar), they are panel specific. If the model is heteroscedastic and variance includes explanatory variables, is it still panel specific?
