如题，rreg回归出现如下结果，怎么看？
Huber iteration 1: maximum difference in weights = .99840405
Huber iteration 2: maximum difference in weights = .63153289
Huber iteration 3: maximum difference in weights = .16006134
Huber iteration 4: maximum difference in weights = .01788925
Biweight iteration 5: maximum difference in weights = .19324695
Biweight iteration 6: maximum difference in weights = .00092911
Robust regression Number of obs = 1000
F( 1, 998) = 1297.98
Prob > F = 0.0000
------------------------------------------------------------------------------
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
x2 | 9.975346 .2768818 36.03 0.000 9.432008 10.51868
_cons | -4.562587 .2807054 -16.25 0.000 -5.113427 -4.011746
------------------------------------------------------------------------------
[rreg] is robust regression, could resist the pull of outliers. The first step is still the OLS, then set aside (downweights) observations that have large residuals. The interpretation is the same as that of [reg]. The key point is when should robust regression be used.
ps. [reg y x1 x2], [rreg y x1 x2..], [reg y x1 x2, vce(robust)] are differnt things
扫码加好友，拉您进群
全部版块
我的主页
收藏
