5:10 p.m.
Hi everybody,
with a little delay I've become aware of some work that updates the good
old area-wide model dataset. In gretl we currently ship AWM, AWM17, and
AWM18, and then it wasn't updated anymore at the source. Now there's
˙Ipek and Kısacıko˘glu who offer new data, see
https://mahmuts.bilkent.edu.tr/ and
https://drive.google.com/file/d/1WRyq4iX3KIEFogV_xESpBtdrT7cQLk3c/view?us....
I wonder whether we should ask them if we can convert their Excel file
and ship it with gretl in the future. The "downside" is that they want
to update it regularly, so the question is whether it's too much of a
moving target; plus I just tested the import into gretl, and it works
flawlessy, you only need to tell gretl to use the second (out of three)
sheets in the workfile. So in that sense it could be argued that there
isn't much value added in providing it in gretl format. (Well, from
their dataset you don't get the series descriptions, so that might be
some value added.)
cheers
sven
7:09 a.m.
On 28/03/2025 18:10, Sven Schreiber wrote:
Hi everybody,
with a little delay I've become aware of some work that updates the good
old area-wide model dataset. In gretl we currently ship AWM, AWM17, and
AWM18, and then it wasn't updated anymore at the source. Now there's
˙Ipek and Kısacıko˘glu who offer new data, see https://
mahmuts.bilkent.edu.tr/ and https://drive.google.com/file/
d/1WRyq4iX3KIEFogV_xESpBtdrT7cQLk3c/view?usp=sharing.
I wonder whether we should ask them if we can convert their Excel file
and ship it with gretl in the future. The "downside" is that they want
to update it regularly, so the question is whether it's too much of a
moving target; plus I just tested the import into gretl, and it works
flawlessy, you only need to tell gretl to use the second (out of three)
sheets in the workfile. So in that sense it could be argued that there
isn't much value added in providing it in gretl format. (Well, from
their dataset you don't get the series descriptions, so that might be
some value added.)
Here is a gdt version of the dataset, complete with variable
descriptions and a placeholder for the dataset info.
However, as the authors explain in the paper, four of the series
(CAN_YEN, NFN_YEN, YWD and YWDX) could not be updated, and only contain
NAs, so I wonder if it's worthwhile to supply those at all. Maybe we
could get in touch with these guys.
-------------------------------------------------------
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
-------------------------------------------------------
10:30 a.m.
Am 31.03.2025 um 09:09 schrieb Riccardo (Jack) Lucchetti:
On 28/03/2025 18:10, Sven Schreiber wrote:
> with a little delay I've become aware of some work that updates the
> good old area-wide model dataset. In gretl we currently ship AWM,
> AWM17, and AWM18, and then it wasn't updated anymore at the source.
> Now there's ˙Ipek and Kısacıko˘glu who offer new data, see https://
> mahmuts.bilkent.edu.tr/ and https://drive.google.com/file/
> d/1WRyq4iX3KIEFogV_xESpBtdrT7cQLk3c/view?usp=sharing.
>
> I wonder whether we should ask them if we can convert their Excel
> file and ship it with gretl in the future. The "downside" is that
> they want to update it regularly, so the question is whether it's too
> much of a moving target; plus I just tested the import into gretl,
> and it works flawlessy, you only need to tell gretl to use the second
> (out of three) sheets in the workfile. So in that sense it could be
> argued that there isn't much value added in providing it in gretl
> format. (Well, from their dataset you don't get the series
> descriptions, so that might be some value added.)
Here is a gdt version of the dataset, complete with variable
descriptions and a placeholder for the dataset info.
However, as the authors explain in the paper, four of the series
(CAN_YEN, NFN_YEN, YWD and YWDX) could not be updated, and only
contain NAs, so I wonder if it's worthwhile to supply those at all.
Maybe we could get in touch with these guys.
Thanks, Jack. I haven't read the details yet, so I wonder why the whole
time span has NAs there for these variables, not just the recent past.
Anyway, by "get in touch with these guys", do you also mean to ask if
they're OK with the distribution, or what (else) did you have in mind?
cheers
sven
>
> -------------------------------------------------------
> 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
> -------------------------------------------------------
>
> _______________________________________________
> Gretl-devel mailing list -- gretl-devel(a)gretlml.univpm.it
> To unsubscribe send an email to gretl-devel-leave(a)gretlml.univpm.it
> Website: https://gretlml.univpm.it/postorius/lists/gretl-devel.gretlml.univpm.it/
10:58 a.m.
On 31/03/2025 12:30, Sven Schreiber wrote:
Am 31.03.2025 um 09:09 schrieb Riccardo (Jack) Lucchetti:
> However, as the authors explain in the paper, four of the series
> (CAN_YEN, NFN_YEN, YWD and YWDX) could not be updated, and only
> contain NAs, so I wonder if it's worthwhile to supply those at all.
> Maybe we could get in touch with these guys.
>
Thanks, Jack. I haven't read the details yet, so I wonder why the whole
time span has NAs there for these variables, not just the recent past.
Anyway, by "get in touch with these guys", do you also mean to ask if
they're OK with the distribution, or what (else) did you have in mind?
Well, since they uploaded the data to a public repo, I can't see what
could be wrong with us repackaging the dataset, provided we cite the
source appropriately.
I was thinking that they may explain to us what in their opinion the
best thing to do with those missing series is, and perhaps we could ask
them if they're willing to provide the data (with all subsequent
updates) in gretl format themselves on their web page, since the work
involved is rather minimal (setting up a gretl script for producing the
gdt file from csv or excel data is trivial).
-------------------------------------------------------
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
-------------------------------------------------------
1:12 p.m.
On 31/03/2025 12:58, Riccardo (Jack) Lucchetti wrote:
> Thanks, Jack. I haven't read the details yet, so I wonder why
the
> whole time span has NAs there for these variables, not just the recent
> past. Anyway, by "get in touch with these guys", do you also mean to
> ask if they're OK with the distribution, or what (else) did you have
> in mind?
Well, since they uploaded the data to a public repo, I can't see what
could be wrong with us repackaging the dataset, provided we cite the
source appropriately.
I was thinking that they may explain to us what in their opinion the
best thing to do with those missing series is, and perhaps we could ask
them if they're willing to provide the data (with all subsequent
updates) in gretl format themselves on their web page, since the work
involved is rather minimal (setting up a gretl script for producing the
gdt file from csv or excel data is trivial).
I just had an email exchange with one of the authors, and we have their
blessing in publishing the dataset with the empty series removed
(they're working to fill them up with new data, so they won't be dropped
in future releases).
I just created a subdir called dataset/AWMD into our private SF
workspace repo with all the necessary files to automate the creation of
the gdt file as much as possible.
I propose to add this dataset to our collection of sample datasets,
possibly moving older versions out of the way.
The original authors have agreed to send us new versions of the dataset
when they release them. We then convert it to the gretl format and
schedule it for inclusion in the next release. Of course, this will mean
that each new gretl version will contain the latest version of the AWMD
available at release time. I believe this shouldn't be a problem, we
already do this with the St. Louis FED database. In the meantime, we
send the authors a copy in gdt format so they can make it available on
their website.
Is this ok for everyone?
-------------------------------------------------------
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
-------------------------------------------------------
2:44 p.m.
On Tue, Apr 1, 2025 at 9:12 AM Riccardo (Jack) Lucchetti
<p002264(a)staff.univpm.it> wrote:
I just had an email exchange with one of the authors, and we have their
blessing in publishing the dataset with the empty series removed
(they're working to fill them up with new data, so they won't be dropped
in future releases).
I just created a subdir called dataset/AWMD into our private SF
workspace repo with all the necessary files to automate the creation of
the gdt file as much as possible.
I propose to add this dataset to our collection of sample datasets,
possibly moving older versions out of the way.
The original authors have agreed to send us new versions of the dataset
when they release them. We then convert it to the gretl format and
schedule it for inclusion in the next release. Of course, this will mean
that each new gretl version will contain the latest version of the AWMD
available at release time. I believe this shouldn't be a problem, we
already do this with the St. Louis FED database. In the meantime, we
send the authors a copy in gdt format so they can make it available on
their website.
Is this ok for everyone?
This all sounds good, Jack!
Allin
4:01 p.m.
On 01/04/2025 16:44, Cottrell, Allin wrote:
> Is this ok for everyone?
This all sounds good, Jack!
OK, I just pushed to git a commit where the new version is added under
the name AWM_update.gdt, the old AWM.gdt and AWM18.gdt have an updated
description and dataset info, and AWM17 was removed from the list of
sample files (the actual gdt file is still there, though, just in case).
-------------------------------------------------------
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
-------------------------------------------------------
11:51 a.m.
On Tue, Apr 1, 2025 at 12:08 PM Riccardo (Jack) Lucchetti
<p002264(a)staff.univpm.it> wrote:
On 01/04/2025 16:44, Cottrell, Allin wrote:
>> Is this ok for everyone?
>
> This all sounds good, Jack!
OK, I just pushed to git a commit where the new version is added under
the name AWM_update.gdt, the old AWM.gdt and AWM18.gdt have an updated
description and dataset info, and AWM17 was removed from the list of
sample files (the actual gdt file is still there, though, just in case).
I'm seeing the updates to AWM.gdt and AWM18.gdt OK, but it seems that
AWM_update.gdt didn't actually get added to the repo.
Allin
12:03 p.m.
On 02/04/2025 13:51, Cottrell, Allin wrote:
I'm seeing the updates to AWM.gdt and AWM18.gdt OK, but it seems
that
AWM_update.gdt didn't actually get added to the repo.
It should be ok now
-------------------------------------------------------
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
-------------------------------------------------------
2
days inactive
7
days old
8 comments
3 participants
participants (3)
-
Cottrell, Allin -
Riccardo (Jack) Lucchetti -
Sven Schreiber