Dear Stata users,

I am working on "the trade potential between China and Central and Eastern European Countries". My dataset is from 2007- 2020, including 17 CEE countries. My output is the total trade volume, and the input includes - GDPi, GDPj, Population(I), Population(j), Distance(ij) and Landlock (dummy variable). The exogenous variables are tariff rate, logistic performance index, government effectiveness, business freedom, labor freedom, trade freedom, EU member, Eurozone, and WTO member. The last three variables are dummy variables.

I tried to use the Battese & Coelli(1995) model, taking the one-step method to parameterize the inputs and exogenous determinants of efficiency at the same time. However, no matter how I adjust my variables, it keeps saying that "could not calculate numerical derivatives -- flat or discontinuous region encountered".

My code is: sfpanel ltrade lgdpc lgdppart lpopc lpoppart ldist landlock year, model(bc95) dist(tn) emean( tariffrate lpi goveffectiveness businessfreedom laborfreedom tradefreedom eum euz wtom ) ort(o)



Attached picture is my code and results.

Array