[Limdep Nlogit List] Specifying panel data structure

[William Greene]

Dear Richard. You do not need to do anything further with the data once
the two variables are in place. For fitting models with two way fixed or
random effects, it will be convenient (and necessary) to realign the time
variable so that the first value is one. You can do that with a trivial
CREATE command. To use the panel data estimators, you need to have a group
size variable, which you can create several ways, most easily
REGRESS;LHS=ONE;RHS=ONE;STR=Firm_ID;Panel$
Thereafter, use ;PDS=_Groupti to invoke any of the panel data estimators.
/Bill Greene

----- Original Message -----
From: "Richard Harris" <r.harris at socsci.gla.ac.uk>
To: limdep at limdep.itls.usyd.edu.au
Sent: Friday, November 27, 2009 1:44:03 PM GMT -05:00 US/Canada Eastern
Subject: [Limdep Nlogit List] Specifying panel data structure

I presume this is a fairly common request. Suppose I have an (unbalanced)
panel comprising a Œpanel¹ variable such as a firm ID marker, and a time
variable (e.g. Year). Thus my database is structured such the first variable
ŒFirm_id¹ takes on unique values for each firm, while the second variable
(Œyear¹) states the year in which the firm is observed. In STATA I would
specify ³tsset firm_id year². What do I do in LIMDEP?

Any help

Thanks

Richard Harris


Prof. Richard Harris
Cairncross Professor of Applied Economics &
Director of CPPR
Ivy Lodge
63 Gibson Street
University of Glasgow G12 8LR
Tel: +44 (0)141 330 4672
Mobile: 07969697224
Fax: +44 (0)141 330 1880
email: Richard.Harris at lbss.gla.ac.uk <mailto:R.Harris at socsci.gla.ac.uk>
http://www.gla.ac.uk/departments/economics/
http://www.cppr.ac.uk <http://www.cppr.ac.uk>

The University of Glasgow, charity number SC004401


_______________________________________________
Limdep site list
Limdep at limdep.itls.usyd.edu.au
http://limdep.itls.usyd.edu.au