I noticed today that a Mata program I developed that calls the -biprobit- procedure and that is reported in http://www.stata-journal.com/article...article=st0423 must be modified to run in Stata v15.
The basic issue is that the way the -athrho- parameter is named in -biprobit- has changed (the change may have occurred in v14, but I jumped immediately from v13 to v15). This code shows the difference
The smaller issue is that anyone wanting to use the method I proposed in the aforelinked Stata Journal paper should add a "version 13" to the top of the code. The larger issue is to alert anyone using -biprobit- to this change.
P.S. I may have missed a general update statement regarding parameter names in v15 (or possibly v14), and apologize if the information provided here is redundant.
The basic issue is that the way the -athrho- parameter is named in -biprobit- has changed (the change may have occurred in v14, but I jumped immediately from v13 to v15). This code shows the difference
Code:
. version 13
. biprobit y1 y2 x
[results suppressed]
. matrix list e(b)
e(b)[1,5]
y1: y1: y2: y2: athrho:
x _cons x _cons _cons
y1 -.08699015 .05623618 .08124519 -.03078105 -.10348424
. version 15
. biprobit y1 y2 x
[results suppressed]
. matrix list e(b)
e(b)[1,5]
y1: y1: y2: y2: /:
x _cons x _cons athrho
y1 -.08699015 .05623618 .08124519 -.03078105 -.10348424
P.S. I may have missed a general update statement regarding parameter names in v15 (or possibly v14), and apologize if the information provided here is redundant.

Comment