Am 29.06.2010 10:46, schrieb Marcus Marktanner: Normally I would start by saying please start a new thread for a new topic, but here by coincidence the solution is probably similar to the one in the thread that you replied to. So something like: genr time matrix greencountries = {} numofunits = max($unit) loop for i=1..numofunits smpl $unit=i --restrict smpl 1961 1980 # not sure if this works here ols yield const time temp1 = $coeff(time) smpl $unit=i --restrict --replace smpl 1981 2000 ols yield const time temp2 = $coeff(time) green = temp2>temp1 ? 1 : 0 matrix greencountries = greencountries | green smpl full endloop Bugs probably included. hth, sven

On Tue, 29 Jun 2010, Sven Schreiber wrote: Sven's idea is fine, but that last "smpl" line will work only for time-series data. Let's suppose you have a variable called "year" in your data set which identifies the year. (If you don't already, you can easily construct one, e.g. "year = time + 1960"). Here's a tested version of Sven's idea which fakes what you want to do (as I understand it), using the panel dataset greene14_1.gdt that's supplied with gretl: <script> open greene14_1.gdt genr time matrix greencountries = {} numofunits = max($unit) loop for i=1..numofunits # first time sub-sample smpl ($unit==i && year>=1970 && year<=1978) --restrict ols Q const time --quiet temp1 = $coeff(time) # second time sub-sample smpl ($unit==i && year>1978) --restrict --replace ols Q const time --quiet temp2 = $coeff(time) green = (temp2 > temp1) greencountries |= green smpl full endloop print greencountries </script> Allin Cottrell

On Tue, 29 Jun 2010, Sven Schreiber wrote: This idea looks quite attractive at first, but I'm afraid it gets confusing fast. The trouble is that we have two modes of setting the sample, which behave somewhat differently and have different support behind the scenes: (a) just shifting the end-points; and (b) carving a chunk out of the full dataset on some criterion, and making the chunk the current dataset for the time being. In a panel dataset in the form of stacked time-series (which is the representation gretl uses) you can sample in mode (a) by choosing a contiguous subset of individuals or units, but sampling by time period necessarily involves mode (b): we have to carve the dataset up; it's a --restrict operation. The syntax you're suggesting in effect "dresses up" a --restrict as if it were a simple moving of end points. But we can't make this sort of sampling behave in that way and I suspect it will end in tears. Allin Cottrell

Dear Allin and Sven, Many thanks about your solution for panel data unit-root test in gretl. It's a must for panel data studies. By the way, the script works perfectly fine here now. Thank you very much. Best regards, Bruno Thiago Tomio -----Mensagem original----- De: gretl-users-bounces(a)lists.wfu.edu [mailto:gretl-users-bounces@lists.wfu.edu] Em nome de Sven Schreiber Enviada em: terça-feira, 29 de junho de 2010 08:47 Para: Gretl list Assunto: Re: [Gretl-users] Re. Panel models tests Am 29.06.2010 01:51, schrieb Allin Cottrell: > Bruno replied, 'I get "Data error" when trying to run tbar = > meanc(tstats)'. I just tried Sven's script and got the same error, > which really should not happen. A work-around is to force the scalar > $test into matrix format, as in > matrix tstats = tstats | {$test} > But now, in gretl CVS and snapshots, Sven's original formulation will > work OK. What's strange is that Bruno's problem turned out to be something else, and then the script worked for him. At least that's how I understood him. Somebody else (Yi-Nung Yang) said he (she?) also ran the script successfully. So I'm a bit confused as to what the bug really was. But it probably doesn't matter, so long as it is running now :-) thanks, sven _______________________________________________ Gretl-users mailing list Gretl-users(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-users