Re: [Gretl-devel] Forecast based on ordered probit
Thank you for all the quick replies.
Am 16.12.2015 um 17:57 schrieb Riccardo (Jack) Lucchetti:
oprobit_predict is by Allin (credit where it's due ;) )
Definitely: Thank you Allin for providing this nice and lean function! ;-)
That said: here's a slight modification of your script:
Thanks for modifying this, Jack. It works nicely.
Now, having sorted the mechanical aspect out, a more interesting
problem
remains open: how do you evaluate forecasts in a case like this? For
continuous outcomes it makes sense to rely on some relatively
straightforward loss functions (also, take a look at Giulio Palomba's
package for the Diebold-Mariano test). But how do you even _define_ the
loss function for a discrete outcome? Hmmm... mumble, mumble...
This is definitely the next start. Actually, I've got Giulio's package
already on the list for this. Also Oleg's reference to the Handbook
article is on the desk for reading between Christmas and New Year :-)
Best,
Artur