悬赏 20 个论坛币 已解决
做三阶段DEA模型,第二阶段随机前沿分析结果如下,请问如何得到投入调整项,做第三阶段。看到有人说冗余项问题,请问从结果中怎么看出冗余项,或者是实际值和最优值。顺便看看结果有没有不合理的地方。求高手指点~~谢谢!
Frontier结果如下:
Output from the program FRONTIER (Version 4.1c)
instruction file = 2.ins
data file = 2.dta
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 -0.63646150E+02 0.24816271E+02 -0.25646944E+01
beta 1 0.97407945E-05 0.69556802E-05 0.14004086E+01
beta 2 0.21305168E+03 0.46685713E+02 0.45635306E+01
sigma-squared 0.41304394E+04
log likelihood function = -0.80787543E+03
the estimates after the grid search were :
beta 0 0.81373526E+01
beta 1 0.97407945E-05
beta 2 0.21305168E+03
sigma-squared 0.91978533E+04
gamma 0.88000000E+00
mu is restricted to be zero
eta 0.00000000E+00
iteration = 0 func evals = 20 llf = -0.76118249E+03
0.81373526E+01 0.97407945E-05 0.21305168E+03 0.91978533E+04 0.88000000E+00
0.00000000E+00
gradient step
iteration = 5 func evals = 81 llf = -0.75058112E+03
0.34458334E+02 0.16068877E-04 0.22512571E+03 0.91978644E+04 0.89695757E+00
0.16392841E-01
iteration = 10 func evals = 195 llf = -0.74883556E+03
0.57074505E+02 0.12818264E-04 0.20263481E+03 0.91982546E+04 0.91471491E+00
0.10198098E-01
iteration = 15 func evals = 325 llf = -0.74577691E+03
0.83276912E+02 0.13370602E-04 0.16526145E+03 0.91988374E+04 0.92302273E+00
0.21311890E-01
iteration = 20 func evals = 446 llf = -0.74427797E+03
0.11769823E+03 0.12276196E-04 0.12003268E+03 0.91995710E+04 0.92491498E+00
0.19897549E-01
iteration = 25 func evals = 580 llf = -0.74402411E+03
0.11880329E+03 0.11529984E-04 0.12424679E+03 0.93825625E+04 0.93085302E+00
0.19466084E-01
iteration = 30 func evals = 721 llf = -0.74312622E+03
0.12273337E+03 0.12418299E-04 0.11243188E+03 0.99955776E+04 0.93915630E+00
0.18940012E-01
iteration = 35 func evals = 848 llf = -0.74306576E+03
0.12661818E+03 0.12484081E-04 0.10748486E+03 0.10078011E+05 0.93975857E+00
0.22435576E-01
iteration = 40 func evals = 999 llf = -0.74286498E+03
0.11708929E+03 0.13661856E-04 0.11213992E+03 0.10351155E+05 0.93594035E+00
0.23636164E-01
iteration = 45 func evals = 1137 llf = -0.74282149E+03
0.11444553E+03 0.13994904E-04 0.11337130E+03 0.10617728E+05 0.93771398E+00
0.23628725E-01
iteration = 50 func evals = 1272 llf = -0.74241455E+03
0.12576177E+03 0.13426278E-04 0.10453725E+03 0.10951494E+05 0.94628576E+00
0.22562590E-01
iteration = 55 func evals = 1416 llf = -0.74159668E+03
0.14374276E+03 0.14741576E-04 0.84672984E+02 0.11766121E+05 0.94850537E+00
0.27013683E-01
iteration = 60 func evals = 1552 llf = -0.74052489E+03
0.16808598E+03 0.14187179E-04 0.70151139E+02 0.12352234E+05 0.95046786E+00
0.22719760E-01
iteration = 65 func evals = 1686 llf = -0.74027910E+03
0.17692242E+03 0.14260752E-04 0.59115544E+02 0.12792201E+05 0.95560080E+00
0.18010423E-01
iteration = 70 func evals = 1831 llf = -0.73975743E+03
0.16308521E+03 0.14653321E-04 0.72198866E+02 0.13115498E+05 0.95857362E+00
0.20521170E-01
iteration = 75 func evals = 1966 llf = -0.73955709E+03
0.16719752E+03 0.14142909E-04 0.68914206E+02 0.13588534E+05 0.96217556E+00
0.19650422E-01
iteration = 80 func evals = 2113 llf = -0.73938524E+03
0.15756876E+03 0.14794863E-04 0.77161216E+02 0.13820303E+05 0.96215850E+00
0.20560286E-01
iteration = 85 func evals = 2257 llf = -0.73893004E+03
0.17131346E+03 0.15661203E-04 0.64102391E+02 0.14243135E+05 0.96117715E+00
0.19804635E-01
iteration = 90 func evals = 2399 llf = -0.73874209E+03
0.18178358E+03 0.15802185E-04 0.54947637E+02 0.14811914E+05 0.96116866E+00
0.20088306E-01
iteration = 95 func evals = 2551 llf = -0.73783939E+03
0.19625227E+03 0.12795459E-04 0.47377018E+02 0.15527030E+05 0.96814148E+00
0.16807640E-01
maximum number of iterations reached
iteration = 100 func evals = 2660 llf = -0.73780420E+03
0.19968629E+03 0.12242235E-04 0.47394116E+02 0.15564183E+05 0.96873619E+00
0.16119161E-01
the final mle estimates are :
coefficient standard-error t-ratio
beta 0 0.19968629E+03 0.25537022E+02 0.78194823E+01
beta 1 0.12242235E-04 0.39160415E-05 0.31261760E+01
beta 2 0.47394116E+02 0.25347647E+02 0.18697639E+01
sigma-squared 0.15564183E+05 0.14685054E+04 0.10598656E+02
gamma 0.96873619E+00 0.57490231E-02 0.16850449E+03
mu is restricted to be zero
eta 0.16119161E-01 0.69924618E-02 0.23052198E+01
log likelihood function = -0.73780537E+03
LR test of the one-sided error = 0.14014013E+03
with number of restrictions = 2
[note that this statistic has a mixed chi-square distribution]
number of iterations = 100
(maximum number of iterations set at : 100)
number of cross-sections = 29
number of time periods = 5
total number of observations = 145
thus there are: 0 obsns not in the panel
covariance matrix :
0.65213949E+03 -0.51640000E-04 -0.57684199E+03 0.25256357E+05 0.79865660E-01
-0.98634725E-02
-0.51640000E-04 0.15335381E-10 0.38912110E-04 0.56286930E-03 -0.50206486E-08
-0.50902253E-08
-0.57684199E+03 0.38912110E-04 0.64250321E+03 -0.23349062E+05 -0.58307642E-01
0.11594707E-01
0.25256357E+05 0.56286930E-03 -0.23349062E+05 0.21565080E+07 0.68402790E+01
-0.47368930E+00
0.79865660E-01 -0.50206486E-08 -0.58307642E-01 0.68402790E+01 0.33051266E-04
-0.47525051E-05
-0.98634725E-02 -0.50902253E-08 0.11594707E-01 -0.47368930E+00 -0.47525051E-05
0.48894522E-04
technical efficiency estimates :
efficiency estimates for year 1 :
firm eff.-est.
1 0.10999430E-01
2 -0.24745802E-01
3 0.45215049E+00
4 0.19543527E+00
5 -0.25676051E-01
6 0.25232248E+00
7 0.28133206E+00
8 -0.25756669E-01
9 0.22961763E-01
10 0.81931282E+00
11 0.53776180E+00
12 0.41340377E+00
13 0.43585622E+00
14 0.88369491E-01
15 -0.25684636E-01
16 0.34577511E-01
17 0.35430336E+00
18 -0.25715443E-01
19 -0.25793932E-01
20 -0.25682679E-01
21 -0.25752774E-01
22 -0.25776489E-01
23 -0.25528096E-01
24 0.75118567E-01
25 0.98630221E+00
26 -0.25160510E-01
27 0.24042315E+00
28 0.33128013E+00
29 -0.25776460E-01
mean eff. in year 1 = 0.18016762E+00
其他时期省略