哪位大侠能帮忙解释下STATA里面的BOXCOX命令具体的计量原理，

　下面有个截图能不能帮忙解释下

. boxcox valueadd capital labor nfirm
Fitting comparison model

Iteration 0: log likelihood = -206.79477
Iteration 1: log likelihood = -201.33464
Iteration 2: log likelihood = -191.34557
Iteration 3: log likelihood = -187.95809
Iteration 4: log likelihood = -187.92461
Iteration 5: log likelihood = -187.92461

Fitting full model

Iteration 0: log likelihood = -151.50376
Iteration 1: log likelihood = -147.47004
Iteration 2: log likelihood = -146.77398
Iteration 3: log likelihood = -146.63832
Iteration 4: log likelihood = -146.63812
Iteration 5: log likelihood = -146.63812

Number of obs = 25
LR chi2(3) = 82.57
Log likelihood = -146.63812 Prob > chi2 = 0.000

------------------------------------------------------------------------------
valueadd | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
/theta | .8461534 .0508321 16.65 0.000 .7465244 .9457825
------------------------------------------------------------------------------

Estimates of scale-variant parameters
----------------------------
| Coef.
-------------+--------------
Notrans |
capital | .7565717
labor | 1.984601
nfirm | .0025507
_cons | 30.36966
-------------+--------------
/sigma | 34.89753
----------------------------

---------------------------------------------------------
Test Restricted LR statistic P-value
H0: log likelihood chi2 Prob > chi2
---------------------------------------------------------
theta = -1 -208.89788 124.52 0.000
theta = 0 -171.75091 50.23 0.000
theta = 1 -151.50376 9.73 0.002
---------------------------------------------------------