yexingtianma 发表于 2011-4-30 23:38 
我遇到了楼主同样的问题 gama值太高,环境变量都不显著,测出的效率值很诡异,
楼主,你从哪里看到环境变量不显著?我都不知道从哪里看到,用Frontier软件做的。下面是我的结果。
the final mle estimates are :
coefficient standard-error t-ratio
beta 0 -0.10692908E-02 0.13794677E+00 -0.77514743E-02
beta 1 -0.77080639E-05 0.95051548E-03 -0.81093512E-02
beta 2 0.11823871E-03 0.13944780E-01 0.84790655E-02
sigma-squared 0.45332487E+00 0.12752648E+02 0.35547508E-01
gamma 0.99999716E+00 0.52680303E-03 0.18982373E+04
mu -0.34178657E+00 0.13495766E+01 -0.25325466E+00
eta is restricted to be zero
log likelihood function = 0.24824985E-01
LR test of the one-sided error = 0.18620315E+02
with number of restrictions = 2
[note that this statistic has a mixed chi-square distribution]