Displaying 1 result from an estimated 1 matches for "vars_".
Did you mean:
vars
2012 Oct 27
0
[gam] [mgcv] Question in integrating a eiker-white "sandwich" VCV estimator into GAM
...t" in the
econometrics sense of the term -- an individual-specific intercept. I
also want to model the continuous variables flexibly -- I have no good
priors on the proper specification for the function form. The model is
the following:
y_{it} = \alpha_i + \beta_1(T_{it}) + f(continuous.vars_{it}) + e_{it}
To control for unobserved time-invariant heterogeneity, I want to
de-mean the data as follows:
y_{it}-\bar{y_i} = \beta_1(T_{it}-\bar{T_i}) +
f(continuous.vars_{it}-\bar{continuous.vars_i}) + e_{it} - \bar{e}_i
Fitting the demeaned model should give me coefficient estimates equal...