#150 fcast --plot command not working

[Tadej Janež]

I'm trying to build a system of equations and use it to forecast different variables.
Here are the commands:
open /home/tadej/Documents/Graduate/markoekonomsko_modeliranje/Data/SLOPOL8.gdt
genr NXR = EXR - IMPR
genr GDPDEF = GDPN / GDPR
genr DIR = GDPR - VAT / GDPDEF
diff GDPR

"slomo" <- system method=tsls
identity GDPR = CR + INVR + GR + NXR
equation CR const DIR
equation INVR const GOV10Y
end system
fcast 2004:1 2008:4 GDPR --plot=test.pdf

I've tried using various --plot variants (e.g. --plot, --plot=test.pdf, --plot=/home/tadej/plot.pdf), but gretl doesn't produce any plot.
Meanwhile, using GUI to produce the forecast using Analysis -> Forecasts -> GDPR produces the plot.

Am I doing something wrong? Or is it a bug in gretl?

Version:
gretl-1.9.3-1.fc13.i686

Thanks for your reply,
Tadej

[Allin Cottrell]

It's a documentation bug, I'd say. The --plot option for the "fcast" command
is valid only for single-equation use. In the GUI you can get a forecast
plot after system estimation, but you have to select a particular variable
to forecast, and there's no facility to make this selection in the context of
the fcast command.

[Sven S.]

Not sure if I understand correctly, but if gretl isn't reacting at all to the 'fcast' command in a particular context (i.e., no error message is given), then I'd say it's more than a documentation bug. (Still no severe bug of course.)

thanks,
sven

[Sven S.]

[This action is part of a little round of maintenance, sorry for the delay.]

I just checked with gretl 1.9.14 and it's properly documented in the command reference and also gives an error in a script. So I'm closing this bug.
thanks,
Sven

[Sven S.]

Hm, actually I'm not sure if I still have the admin permissions for the bug tracker, I cannot see any way to actually close the ticket...
-sven