|
boxcox y x1 x2
Fitting comparison model
Iteration 0: log likelihood = -125.91879
Iteration 1: log likelihood = -116.28801
Iteration 2: log likelihood = -115.67671
Iteration 3: log likelihood = -115.6754
Iteration 4: log likelihood = -115.6754
Fitting full model
Iteration 0: log likelihood = -112.08395
Iteration 1: log likelihood = -109.39327
Iteration 2: log likelihood = -105.17592
Iteration 3: log likelihood = -103.34288
Iteration 4: log likelihood = -103.03614
Iteration 5: log likelihood = -103.03572
Iteration 6: log likelihood = -103.03572
Number of obs = 15
LR chi2(2) = 25.28
Log likelihood = -103.03572 Prob > chi2 = 0.000
------------------------------------------------------------------------------
y | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
/theta | .4688845 .1068079 4.39 0.000 .259545 .6782241
------------------------------------------------------------------------------
Estimates of scale-variant parameters
----------------------------
| Coef.
-------------+--------------
Notrans |
x1 | .0497033
x2 | .0135919
_cons | 6.813018
-------------+--------------
/sigma | 11.12579
----------------------------
---------------------------------------------------------
Test Restricted LR statistic P-value
H0: log likelihood chi2 Prob > chi2
---------------------------------------------------------
theta = -1 -154.97569 103.88 0.000
theta = 0 -111.16542 16.26 0.000
theta = 1 -112.08395 18.10 0.000
---------------------------------------------------------
.
这是stata的运行结果,与sas和spss数据有所差别,但与sas接近,与spss差别较大,在sas和spss中用lanmda表示,STATA用theta表示
|