defbundle parsing
by Sven S

Hi,
the following line works, but shouldn't it fail due to the trailing comma?:
bundle b = defbundle("a",1 ,)
thanks
sven
1 month, 3 weeks

$xtxinv available after restricted VECM or not?
by Sven Schreiber

Hi,
the function reference says that after restricting the alphas in a VECM
you don't get $xtxinv, and I can see the reason why (I think).
However, this seems to work:
<hansl>
open denmark
vecm 4 1 LRM LRY IBO
restrict
a[1] = 0
b[1] = 1
end restrict --full
eval $coeff
eval $stderr
eval $xtxinv
</hansl>
Bug or feature?
thanks
sven
2 months, 1 week

Probably a bug in stringify()
by Artur Tarassow

Hi all,
I guess I found a bug in the stringify() function. In the following
example both "x" and "z" are constant integer series taking, for
instance, value 1. In case the length of the string array "S" equals
nobs(series), the mapping is _not_ right. However, in case nelem(S) >
nobs(series) things work ok.
<hansl>
set verbose off
nulldata 3
series y = seq(1,3)'
series x = 1
series z = 1
strings S1 = defarray("A", "B", "C")
strings S2 = defarray("A", "B", "C", "D")
scalar err1 = stringify(y, S1)
scalar err2 = stringify(x, S1)
scalar err3 = stringify(z, S2)
print err1 err2 err3
print -o
</hansl>
The last print command yields:
<output>
index y x z
1 1 A A A
2 2 B B A
3 3 C C A
</output>
Obviously, the mapping is not right for x but for z.
This is with 2-days old gretl (git) on Ubuntu 20.04.
Best,
Artur
4 months

$test accessor after biprobit
by Sven Schreiber

Hi,
prompted by Jack's recent biprobit explanation in the guide I checked
whether the automatically run test is available via $test -- it's not
(gives NA).
I know biprobit is not a test but an estimation command. Does that mean
that $test/$pvalue shouldn't be used there? Should there be a new
'modtest' incarnation?
thanks
sven
4 months

move "univariate" cointegration section in the guide
by Sven Schreiber

Hi,
this is not new (I guess), but I noticed section 31.4 about
"Cointegration tests" in the chapter "Univariate time series models". I
find this unfotunate because how can _co_integration be univariate? I
know that sometimes people refer to a single-equation setup as
univariate but I always thought that was at least sloppy and actually wrong.
So I'd like to move that section into the "Cointegration and VECM"
chapter. OK?
thanks
sven
4 months

matrix-to-scalar details
by Sven Schreiber

Hi,
I stumbled over some details of gretl's automatic conversion from
1-element matrices to scalars. See:
<hansl>
m = {3.3}
eval typeof(m) # 3 (matrix)
eval typeof(m[1]) # 1 (scalar)
eval typeof(m[1,]) # 1 (scalar)
</hansl>
I find the last result a bit inconvenient; the "1," specifies an entire
row IMHO, and it's just by coincidence that it only has one element. The
one-element property also isn't the relevant criterion as can be seen
from the first result on the "full" one-element matrix m. The
implication is that doing nelem(m[1,]) fails in this case because of a
wrong type (scalar).
So either it would be good if a row or column-indexing specification
returned a matrix type, or alternatively if nelem() worked on a scalar
as well (and would then return the value 1).
This is with 2020d.
thanks
sven
4 months

Forecasting techniques
by Riccardo (Jack) Lucchetti

Folks, I've just come across this:
https://www.dropbox.com/s/6k4yhuxnmnxu1e6/Forecasting%20theory%20and%20pr...
which contains a nice "bird's eye" introduction to everything that has to
do with economic forecasting.
It's a shame that gretl is never mentioned in the article, but OTOH is
good to see that we support, at least partly, most of the techniques
mentioned here. I'm sharing it with you because it is possible that we may
find some ideas on additions or enhancements to what we currently have.
-------------------------------------------------------
Riccardo (Jack) Lucchetti
Dipartimento di Scienze Economiche e Sociali (DiSES)
Università Politecnica delle Marche
(formerly known as Università di Ancona)
r.lucchetti(a)univpm.it
http://www2.econ.univpm.it/servizi/hpp/lucchetti
-------------------------------------------------------
4 months

keyboard shortcut issues
by Sven Schreiber

Hi,
I learned something new about gretl's interface even after all these
years by looking in the help menu at the keyboard shortcuts page. (I'm
not entirely sure it has to be a PDF page with lots of whitespace
margins, but that's not the point right now.)
'h' (open command reference) seems to outdated or broken at least on
Windows. Instead the menu indicates ctrl-H which works alright, although
it's actually ctrl-h (lowercase). Should the shortcuts help page be updated?
thanks
sven
4 months

new operation on string-valued series
by Allin Cottrell

I'm not documenting this yet because it needs more testing, but in
current git I've enabled a new operation on string-valued series:
logical product. (Snapshots to follow.)
If sv1 and sv2 are string-valued series, then
series sv3 = sv1 * sv2
now yields another string-valued series with value si_sj at
observations where sv1 has value si and sv2 has value sj.
A simple example is afforded by the R dataset warpbreaks (csv
version attached). This includes two "factor" series, wool (with
values "A" and "B") and tension (with values "L", "M" and "H").
If you multiply them together you get a series with 6 distinct
values, "A_L" to "B_H".
<hansl>
open warpbreaks.csv
series wt = wool*tension
print -o
</hansl>
Allin
4 months, 2 weeks

feedback on the new tdisagg feature
by Sven Schreiber

Hi Allin and everybody who's interested,
here are some first comments about the new tdisagg function, as
described in chapter 9 of the current version (in git) of the gretl user
guide:
- mention Di Fonzo's affiliation
- I don't find the money stock argument entirely convincing: One may
well want to take the average stocks in a quarter instead of only the
final value.
- In 9.2 from a gretl object point of view I find it confusing to talk
about a matrix holding a series. Maybe it would be good already at that
point to clarify that the tdisagg function accepts both series or matrix
arguments. Similarly in section 9.8 it says Y can be a list of g series,
but the function ref says for Y only "series or matrix" (and not list).
(Also at the top of 9.3, but then otherwise further down.)
- Section 9.9 / Examples: Not clear to me why examples are online
(only). Isn't the standard way in gretl to ship the example scripts?
- _() is used but is still not documented (I think)
- I think it would make sense to have defaults for s. For example, if we
are in the context of a quarterly dataset, I think it's natural to have
s=3. And so on, so a $pd-dependent default. (Obviously, this won't be
workable in all cases, but in some.)
- It should be mentioned that chowlin existed before and is now superseded.
- I like the explanation how to assign to a series!
- Why is in 9.4 Y now called Y0, while X is still X?
- In 9.4 I'd suggest to give the string options in double quotes, since
that's how they must appear in the bundle definition.
- Doesn't the result bundle have to be passed in pointer form? Otherwise
I don't see how it can be altered in hansl. Well I see that it actually
works (only) in non-pointer form, but I'm puzzled -- isn't this against
the hansl syntax spec??
thanks
Sven
4 months, 3 weeks