Panel unit root test for unbalanced panel

Questions related to panel (pooled cross-section time series) data.
lis
Posts: 1
Joined: Wed Mar 25, 2009 5:29 am

Panel unit root test for unbalanced panel

Unread post by lis »

I am a new rats user. I am trying to run the "RATS CODE FOR COINTEGRATION TESTS IN HETEROGENEOUS PANELS WITH MULTIPLE REGRESSORS" by Pedroni for an unbalanced panel. My code looks as follows below (see part of the code). The problem is now that I have a line break and obtain the following syntax error when I run Rats:
Normal Completion. Halt at XXX
## SX16. Missing Operand or Adjacent Operators
>>>>23,23,23,23,23,23,<<<<

Who can help mw out on this issue?
Thanx a lot!
Cheers,
Lisa

Part of the code:

compute Nsecs = 138 ;* cross section dimension, N
compute Tperiods = 23 ;* time series dimension, T, if balanced
compute mlag = 0 ;* largest possible value for lag truncations (0,1,2; Isa: 1-4))
compute yestrend = 0 ;* set to 1 for hetero trends, else 0
compute m = 0 ;* number of RHS regressors (2 variablen)

dec vec[series] datavec(m+1) ;* define data variables below
set datavec(1) = budget ;* LHS variable
* set datavec(2) = FFII ;* second RHS variable, and so forth
* set datavec(3) = SER ;* second RHS variable, and so forth

compute unbal = 1 ;* set to 1 for unbalanced panel, else 0
dec vec Tvec(Nsecs)

* For unbalanced panels, fill in the value below for T for each member
* in the order in which they appear in the stacked panel
* compute Tvec = ||100,100,100 ... one value for each member .... 100, ||

compute Tvec= || 23,23,23,23,23,23,14,23,23,23,23,23,23,23,23,23,23,23,23,21,23,23,23,23,23,23,23,23,23,23,23,23,23,23,23,23,23,23,23,18,
23,23,23,23,23,23,23,23,23,23,23,23,23,23,20,23,23,23,23,23,23,23,23,20,23,23,23,23,23,23,23,23,23,
|23,23,23,23,16,23,23,23,23,23,23,23,23,23,23,23,23,23,23,23,23,23,23,23,23,23,23,23,16,23,23,23,
23,23,23,23,23,23,23,23,23,23,23,23,23,23,23,19,23,23,23,23,21,20,23,19,23,23,23,22,23,23,20,23,23,||
TomDoan
Posts: 7814
Joined: Wed Nov 01, 2006 4:36 pm

Re: Panel unit root test for unbalanced panel

Unread post by TomDoan »

You're using a rather old version of this that requires a lot of editing. Try using the newer @PanCoint procedure. That's on the distribution CD as well.
Post Reply