[Gretl-users] Re: A query re the equation version of quantile regression.
Hi
it is not the 0,5 quantile equation as that has been one estimated. The
result seems to lay between 0.1 and 0.2 quantiles
On another matter - Basic Structural Model - error message saying the
forecast function will not work with exogenous variables (even though
values have been supplied for the forecast horizon)
Second error -editing graphs in the ARIMA modelling option - the graph
options for fonts, titles, colours, line styles can all be selected.
However on issuing the Apply command -the following error message is given
"C:\Program Files\gretl\wgnuplot.exe"
"C:\Users\Brian\AppData\Roaming\gretl\gpttmp.7XAUW0": exit code 1
Brian
On Mon, 11 Jan 2021 at 15:39, Sven Schreiber <svetosch(a)gmx.net> wrote:
Am 11.01.2021 um 16:01 schrieb Brian Revell:
> It is possible to save an equation version in the quantile regression
> option of Robust methods modelling. However, there is nothing in the
> User guide to indicate which quantile it relates to.
Well, the equation view is really barebones, it doesn't even mention
what kind of model it is in the first place. In my view even the
information there about the residual (abs.) sum could be removed, don't
know why that one was chosen to be shown.
But the quantile is chosen by the user at the specification stage.
Default is the median (0.5). This is shown as tau in the standard model
output.
cheers
sven
_______________________________________________
Gretl-users mailing list -- gretl-users(a)gretlml.univpm.it
To unsubscribe send an email to gretl-users-leave(a)gretlml.univpm.it
Website:
https://gretlml.univpm.it/postorius/lists/gretl-users.gretlml.univpm.it/
Attachments:
- attachment.htm (text/html — 3.2 KB)