[Gretl-users] Re: [EXTERNAL] Re: panplot question
On Fri, Jan 27, 2023 at 11:33 AM Summers, Peter <psummers(a)highpoint.edu> wrote:
Artur & Sven,
Thanks for your help! I'm using the Penn World Tables. Gretl correctly recognized it
as a panel, but I hadn't set the dataset structure. When I did that, it
worked...almost. Running the attached script, the first plot still has observation
numbers, but the rest correctly show the year.
It's the double sub-sampling for the first plot (year >= 1960
--restrict and a short list of countries) that's throwing gretl's
calendar. The subsequent plots are not restricted by time, only by
country (since you use --restrict --replace with smpl). The first plot
works if you use the panel-specific means of sampling in the time
dimension:
instead of: smpl year >= 1960 --restrict
do: smpl 1960 2019 --time
Allin