Re: st: seqlogit error: "No path in the tree leads to values .."

[Maarten Buis <maartenlbuis@gmail.com>]

On Fri, Aug 12, 2011 at 4:06 PM, Rijo John wrote:
> I am trying to estimate a sequential logit model using the seqlogit
> proceedure by Maarten. Here is the command I am executing
>
> seqlogit choice age agesq, tree(1: 2 3 4, 4:2 3)
>
> I keep getting the error message "No path in the tree leads to values 2 3"

-seqlogit- requires that each value of choice can be reached by one
and only one path specified in the -tree()- option. In your case the
value 1 can be reached by failing the first transition, the value 4
can be reached by passing the first and failing the second transition,
but that leaves the values 2 and 3. So, you need to add an additional
transition. Valid specifications of the -tree()- option would for
example be:

tree(1: 2 3 4, 4:2 3, 2:3 )

or:

 tree(1: 2 3 4, 4:2: 3)

> My dependent variable choice is coded with 4 outcomes. Each id in the
> data only chose one of those outcomes. There are 17,109 ids in my data
> and it has been expanded by 4 to a total of 17,109X4

I do not think you want to do that. -seqlogit- is designed for dealing
with the case where there is only one observation per id, the fact
that there are multiple transitions is taken care of in the likelihood
function.

Hope this helps,
Maarten

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany


http://www.maartenbuis.nl
--------------------------
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/