我是利用stata 做double hurdle模型，参考的操作步骤是 The Stata Journal (2013) 13, Number 4, pp. 776–794， ‘Implementation of a double-hurdle model’，数据量并不大，尝试了几次，用不同的自变量，但都是提示initial values not feasible，r(1400)   。
其中r（1400）的具体解释如下：

[P]     error . . . . . . . . . . . . . . . . . . . . . . . . Return code 1400
        numerical overflow;
        You have attempted something that, in the midst of the
        necessary calculations, has resulted in something too large
        for Stata to deal with accurately.  Most commonly, this is
        an attempt to estimate a model (say with regress) with more
        than 2,147,483,647 effective observations.  This effective
        number could be reached with far fewer observations if you
        were running a frequency-weighted model.

貌似是说我的数据太大了吗？但是我的样本量并不大啊，不知道为什么会出现这种情况，有哪位好心人帮忙解释一下吗？不胜感激。