stata做mvprobit多方程的多元probit模型回归出错

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

命令如下：mvprobit (organization_in = size1 size2 lnage lnexp hum rd comp1 comp2 tgovern) (process_in = size1 size2 lnage lnexp hum rd comp1 comp2 tgovern) (product_in = size1 size2 lnage lnexp hum rd comp1 comp2 tgovern) (market_in = size1 size2 lnage lnexp hum rd comp1 comp2 tgovern), nolog dr(100)出现如下错误：could not calculate numerical derivatives -- flat or discontinuous region 求助大神该怎么办？

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：Probit Stata tata Rob bit 模型

试试底下连结是否有帮助？http://www.stata-journal.com/article.html?article=st0423

或 http://www.stata-journal.com/sjpdf.html?articlenum=st0045

黃河泉 发表于 2016-12-15 06:54
试试底下连结是否有帮助？http://www.stata-journal.com/article.html?article=st0423

之前就是参考的这两篇文章的方法，但是不知到为什么，有些特定的变量组回归没有问题，但是换掉一些变量就会出现上述的报错。

Revo1ut1on 发表于 2016-12-15 13:00
之前就是参考的这两篇文章的方法，但是不知到为什么，有些特定的变量组回归没有问题，但是换掉一些变量就 ...

It happens! 这种高维度概似函数（likelihood function）的极大过程中，有些函数"长得""太平"了，以致于无法往前（极大值）迈进！

黃河泉 发表于 2016-12-15 15:04
It happens! 这种高维度概似函数（likelihood function）的极大过程中，有些函数"长得""太平"了，以致于无 ...

你的意思我不太懂诶
这是迭代过程
Iteration 0:   log likelihood = -2619.1934  
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Iteration 1:   log likelihood = -2537.4064  
Iteration 2:   log likelihood = -2312.4194  
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
Warning: cannot do Cholesky factorization of rho matrix
numerical derivatives are approximate
flat or discontinuous region encountered
Warning: cannot do Cholesky factorization of rho matrix
进行迭代之后就会这样报错，一直出现这个无法继续迭代。这个到底是什么原因呢？要怎么处理呢？

黃河泉 发表于 2016-12-15 15:04
It happens! 这种高维度概似函数（likelihood function）的极大过程中，有些函数"长得""太平"了，以致于无 ...

我在原来基础上减少了draw的数值，越小这种情况就出现的越少，但是draw数值越小就越不接近最大似然估计的结果

Revo1ut1on 发表于 2016-12-15 18:58
你的意思我不太懂诶
这是迭代过程
Iteration 0:   log likelihood = -2619.1934

不好意思，恐怕我也无法帮上忙！先试试简单的情况（少一点解释变量！）

黃河泉 发表于 2016-12-16 07:28
不好意思，恐怕我也无法帮上忙！先试试简单的情况（少一点解释变量！）

谢谢您，我自己再试试，您已经帮到我了，您给的第二篇文章的方法我会尝试的。

你好！楼主！请问最后找到解决办法了吗？我现在也遇到了这样的问题！