[编程问题求助] stata做sfa分析的详细命令语句 [推广有奖]

0关注
4粉丝

大专生

61%

还不是VIP/贵宾

威望: 0 级
论坛币: 16 个
通用积分: 19.0456
学术水平: 0 点
热心指数: 0 点
信用等级: 0 点
经验: 176 点
帖子: 21
精华: 0
在线时间: 44 小时
注册时间: 2014-9-27
最后登录: 2017-2-28

楼主

信自己傲天下 发表于 2015-12-20 18:55:31 |只看作者 |坛友微信交流群|倒序 |AI写论文

是否 +2 论坛币

k人参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

stata做sfa分析的详细命令语句

各位群友你们好，就是我在用stata计算sfa的时候，无论是使用frontier+变量，xtfrontier+变量，还是sfcross,sfpanel计算的时候，计算出了的效率值怎么是大于1的，我的一组数据是这样的

lnoutput	lnlabour	lncapital
15.70039162	9.579003174	12.74458941
15.31221415	9.549238236	12.51381099
12.89867818	7.383989458	10.71062391
12.92488172	8.009363077	10.96025997
15.03537952	8.976135873	12.14255227
13.93401911	8.685246776	11.58193927
13.02042633	7.568895663	10.89511208
13.38595366	7.845807503	10.90834761
12.19263983	6.439350371	9.454524999

命令如下：
sfcross  lnoutput lnlabour lncapital
initial:    Log likelihood = -13.449518
Iteration 0: Log likelihood = -13.449518  (not concave)
Iteration 1: Log likelihood = -7.0344072  (not concave)
Iteration 2: Log likelihood = -4.0625023
Iteration 3: Log likelihood =  -3.208996  (not concave)
Iteration 4: Log likelihood = -2.4905345  (not concave)
Iteration 5: Log likelihood = -2.3886018  (not concave)
Iteration 6: Log likelihood = -2.3765454
Iteration 7: Log likelihood = -2.1444908
Iteration 8: Log likelihood =  -2.008752
Iteration 9: Log likelihood = -1.9937579
Iteration 10:  Log likelihood =  -1.986255
Iteration 11:  Log likelihood = -1.9849598
Iteration 12:  Log likelihood = -1.9840791
Iteration 13:  Log likelihood = -1.9839049
Iteration 14:  Log likelihood = -1.9837359
Iteration 15:  Log likelihood = -1.9837269
Iteration 16:  Log likelihood = -1.9836615
Iteration 17:  Log likelihood = -1.9836566
Iteration 18:  Log likelihood = -1.9836439
Iteration 19:  Log likelihood = -1.9836424
Iteration 20:  Log likelihood = -1.9836418
Iteration 21:  Log likelihood = -1.9836416

Stoc. frontier normal/exponential model             Number of obs =       9
                                                   Wald chi2(2)  = 127.03
                                                   Prob > chi2 = 0.0000

Log likelihood = -1.9836
------------------------------------------------------------------------------
      lno |    Coef. Std. Err.    z P>|z|    [95% Conf. Interval]
-------------+----------------------------------------------------------------
Frontier    |
      lnl | .185529 .6727565    0.28 0.783 -1.133049 1.504107
      lnk | .9766153 .6863974    1.42 0.155 -.3686988 2.321929
   _cons | 1.24623 2.708034    0.46 0.645 -4.061419 6.553879
-------------+----------------------------------------------------------------
Usigma    |
   _cons |  -9.548672 264.708 -0.04 0.971 -528.3668 509.2694
-------------+----------------------------------------------------------------
Vsigma    |
   _cons | -2.39786 .5146479 -4.66 0.000 -3.406551 -1.389169
-------------+----------------------------------------------------------------
   sigma_u | .0084437 1.117556    0.01 0.994    1.8e-115 3.9e+110
   sigma_v | .3015167 .0775875    3.89 0.000    .1820861    .499282
   lambda | .028004 1.150761    0.02 0.981 -2.227445 2.283453
------------------------------------------------------------------------------

. predict te1
(option xb assumed; fitted values)

. list te1

   +----------+
   |    te1 |
   |----------|
  1. | 15.46997 |
  2. | 15.23907 |
  3. | 13.07633 |
  4. | 13.43616 |
  5. | 14.77017 |
   |----------|
  6. | 14.16869 |
  7. | 13.29081 |
  8. | 13.35511 |
  9. | 11.67435 |
   +----------+

我想问一下这个te是效率值吗？如果是的话为什么都这么大，还有就是哪些回归系数的显著性判断的那个“*”都是自己查表标上去的，还是说软件会自动输出来，急求懂的人帮忙解释一下  麻烦给一下详细的命令好吗？急求谢谢