原始数据是40个obs

logistic y x1 x2 x3

note: x3 != 1 predicts failure perfectly
         x3 dropped and 7 obs not used

只得到x1 x2的分析结果。

------------------------------------------------------------------------------
           y | Odds Ratio   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
          x1 |   1.835185   1.434904     0.78   0.437     .3964035    8.496146
          x2 |   .9724318   .0311901    -0.87   0.383     .9131821    1.035526
------------------------------------------------------------------------------

为什么没有使用x3呢？