技术效率
Output from the program FRONTIER (Version 4.1c)

instruction file = Eg1-ins.txt
data file =        TE3.txt     

Error Components Frontier (see B&C 1992)
The model is a production function
The dependent variable is not logged

the ols estimates are :

                 coefficient     standard-error    t-ratio

  beta 0         0.10896200E+04  0.88805413E+03  0.12269748E+01
  beta 1        -0.57878362E+00  0.68381598E+00 -0.84640259E+00
  beta 2        -0.12454457E+01  0.11177404E+01 -0.11142530E+01
  beta 3         0.47813449E+00  0.52464828E+00  0.91134290E+00
  sigma-squared  0.17958524E+06

log likelihood function =  -0.41513010E+02

the estimates after the grid search were :

  beta 0         0.11339834E+04
  beta 1        -0.57878362E+00
  beta 2        -0.12454457E+01
  beta 3         0.47813449E+00
  sigma-squared  0.61829851E+05
  gamma          0.50000000E-01
   mu is restricted to be zero
   eta is restricted to be zero


iteration =     0  func evals =     20  llf = -0.41513349E+02
     0.11339834E+04-0.57878362E+00-0.12454457E+01 0.47813449E+00 0.61829851E+05
     0.50000000E-01
gradient step
iteration =     5  func evals =     82  llf = -0.41513309E+02
     0.11339834E+04-0.57991533E+00-0.12485704E+01 0.47730606E+00 0.61829851E+05
     0.44364142E-01
search failed. fn val indep of search direction
iteration =     7  func evals =    104  llf = -0.41513309E+02
     0.11339834E+04-0.57991776E+00-0.12485789E+01 0.47730242E+00 0.61829851E+05
     0.44349681E-01


the final mle estimates are :

                 coefficient     standard-error    t-ratio

  beta 0         0.11339834E+04  0.10362449E+01  0.10943198E+04
  beta 1        -0.57991776E+00  0.32229849E+00 -0.17993189E+01
  beta 2        -0.12485789E+01  0.43995531E+00 -0.28379676E+01
  beta 3         0.47730242E+00  0.28741526E+00  0.16606718E+01
  sigma-squared  0.61829851E+05  0.10000282E+01  0.61828110E+05
  gamma          0.44349681E-01  0.63985277E+00  0.69312322E-01
   mu is restricted to be zero
   eta is restricted to be zero

log likelihood function =  -0.41513309E+02

the likelihood value is less than that obtained
using ols! - try again using different starting values

number of iterations =      7

(maximum number of iterations set at :   100)

number of cross-sections =      6

number of time periods =      1

total number of observations =      6

thus there are:      0  obsns not in the panel

covariance matrix :

  0.10738035E+01  0.28413711E-01  0.92468437E-01  0.38784201E-01  0.20386186E-02
  0.15650389E+00
  0.28413711E-01  0.10387632E+00  0.10608367E-01  0.43427609E-02  0.79823824E-03
  0.11356467E+00
  0.92468437E-01  0.10608367E-01  0.19356068E+00 -0.12026620E-01  0.25799244E-02
  0.21479862E+00
  0.38784201E-01  0.43427609E-02 -0.12026620E-01  0.82607534E-01  0.10778617E-02
  0.39749447E-01
  0.20386186E-02  0.79823824E-03  0.25799244E-02  0.10778617E-02  0.10000563E+01
  0.43136661E-02
  0.15650389E+00  0.11356467E+00  0.21479862E+00  0.39749447E-01  0.43136661E-02
  0.40941156E+00

technical efficiency estimates :

     firm             eff.-est.

       1           0.10000000E+01
       2           0.85785881E+00
       3          -0.15568379E+02
       4           0.87127651E+00
       5           0.95349411E+00
       6           0.88901789E+00


mean efficiency =  -0.18327887E+01