observation labels
by Sven Schreiber
Hi,
about the possibility to add observation labels (or "markers"?) some
practical issues.
- why no "help" button in the dialog ("add some from file now?"), in
analogy to the help button that you get for adding series labels?
- IMHO it would be good to have a facility to type them manually in
gretl. Often it's just 10 or 20 or so countries or regions or whatever.
Right now I had to create this text file with the labels just to satisfy
gretl.
thanks,
sven
9 years
2017b release
by Allin Cottrell
Hello all,
I'd like to release gretl 2017b before the upcoming gretl conference
( http://www.gretlconference.org/ ) which is now not far off.
So this would be a good time to press bug reports or update
translations.
Allin
9 years
fcstats upgrade
by Allin Cottrell
I've recently had cause to hit on gretl's fcstats function and I
found a couple of limitations which are now fixed in git and
snapshots.
For anyone who hasn't used this function (recently), it takes a
series or vector argument y (observed values) and a series or vector
argument f (forecast values) and returns a fairly comprehensive set
of forecast evaluation statistics in the form of a column vector.
Problem 1: the "series or vector" flexibility was not quite what it
might seem, since you had to supply either two series or two
vectors; you couldn't mix and match.
Fix 1: allow mix and match, provided the length of the vector
matches the number of observations in the current sample range (for
a series argument).
Problem 2: OK, but I'd like to provide several columns of forecast
values -- can't you give me a return matrix with several columns of
forecast evaluation statistics?
Fix 2: We now support this. For the actual observed values, y, we
still require a single series or vector, but if you give as the
second argument, f, a matrix with n columns, we return a matrix with
n columns, each of which contains forecast evaluation statistics for
the corresponding column of f. (The number of rows of f must, of
course, match the number of values in y).
The doc for fcstats() will be updated shortly.
Allin
9 years
minuscule doc issue
by Sven Schreiber
Hi,
in section 15.5 the single equality '=' is listed as a matrix test for
equality, I guess it should be updated to reflect the new policy.
cheers,
sven
9 years
new variables created with list command do not appear
by Sven Schreiber
Hi,
here's the minor problem, experienced in the GUI of 2016d and 2017a:
- open the bjg.gdt workfile
- choose "add/ periodic dummies" from the menu (don't know if this is
necessary to trigger the bug)
- the seasonal dummies dm1..dm12 appear in the workfile
- choose "add/ define new variable"
- enter into the dialog "csdum = seasonals(1, 1)", hit OK
- nothing seems to happen, no new variables in the workfile
It turns out that the list 'csdum' and the variables S1..S12 are present
and can be chosen for example in the model specification window. But
they are not visible in the main window.
thanks,
sven
9 years
construct lists in the GUI
by Sven Schreiber
Hi,
so I've started looking at the GUI with a fresh perspective (due to
teaching, of course...).
I think it would be good if there were a (semi-) GUI way of constructing
new lists in a workfile, in analogy to the "define new variable" menu
entry which gives you a one-line text window where you can insert a formula.
Currently we have "Data/Define or edit list", but this only works with a
set of existing series. And of course there's the console.
So my suggestion is to either extend the window that you get when
choosing "Data/Define or edit list", or --perhaps more fitting-- have a
new menu entry "Add/Create new series collection in list" (label too
long, I know...).
The user would get a very simple one-line window, where for example she
could enter "sdums = seasonals(1,1)". The help button might lead to an
enumeration of list-returning functions, but a very generic help would
also be OK (as in "read ch. 14.1 of the guide").
What do you think?
thanks,
sven
9 years
menus AR(1) / autoregr. / vs. "ordinary" AR
by Sven Schreiber
Hi,
although I do a lot of time series, I don't usually use the methods that
come under the heading of "time-series AR(1)" or "autoregressive" (that
is Cochrane-Orcutt etc.). Indeed I would claim that those are in general
not much used anymore -- either people use ARIMA(X) estimators (with AR
as special cases), or they estimate with autocorrelation and then apply
Newey-West standard errors, or they follow the dictum of Wooldridge and
others and try to introduce more regressors to capture the dynamics.
Having this as a background, I have two questions:
First, is it really necessary to have two separate menu entries in the
time-series submenu (of Model) for AR(1) and Autoregressive? AFAICS it's
the same stuff, and that submenu is pretty long already.
Secondly, I'd suggest to put the keyword "errors" or "innovations" in
those menu labels, because for me an AR(1) model is just "ols y const
y(-1)", not the other stuff.
So, instead of "AR(1)" and "Autoregressive estimation" I'd have "With AR
errors".
Thanks,
Sven
9 years
another ternary woe
by Sven Schreiber
Hi,
very short minimal script here:
<hansl>
matrix heyho = ("a" == "b") ? garbl[3,] : I(2)
</hansl>
There is no object "garbl", but that's by design in the case that the
condition isn't true. Gretl says "[" would be invalid here.
thanks,
sven
9 years