7:45 a.m.
? List04 = wes(Cement, 0, null, 1, 0.4, 0.03, 0.3, 2, 0)
Index value -1 is out of bounds
*** error within loop in function wes
genr f[j+1]=(F[j] + m*T[j])*S[j+m-$pd]
I am getting the above error while trying to estimate the holt-winters from fcModels. Any
help?
https://www.dropbox.com/scl/fi/nklcilb87qdtc2dm1klr6/aus_production_cemen...
2:33 p.m.
Hi,
the author of the contributed package fcModels is Yi-Nung Yang. Not sure
if he's still following this list - Yi-Nung, please make yourself heard
if you're here.
Having said that, I'm also getting similar errors when running the wes
function from that package on the djclose data that's used in the
package's sample script. It looks like a bug that perhaps has been there
quite a while already. Looking at the code a bit and the argument label
"number of years...", my suspicion is that the wes function is not
really set up for daily datasets like djclose. Perhaps only annual,
quarterly, and monthly periodicities are supported. Are you also working
with daily data?
Note that there are also two equally established packages "Holt" and
"Winters" by gretl veteran Ignacio (Diaz-Emparanza).
cheers
sven
Am 03.03.2026 um 08:45 schrieb ramakrishnasalagrama(a)gmail.com:
> ? List04 = wes(Cement, 0, null, 1, 0.4, 0.03, 0.3, 2, 0)
> Index value -1 is out of bounds
> *** error within loop in function wes
>> genr f[j+1]=(F[j] + m*T[j])*S[j+m-$pd]
>
> I am getting the above error while trying to estimate the holt-winters from fcModels.
Any help?
>
>
https://www.dropbox.com/scl/fi/nklcilb87qdtc2dm1klr6/aus_production_cemen...
> _______________________________________________
> 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/
4:59 p.m.
On Tue, Mar 3, 2026 at 2:45 AM <ramakrishnasalagrama(a)gmail.com> wrote:
? List04 = wes(Cement, 0, null, 1, 0.4, 0.03, 0.3, 2, 0)
Index value -1 is out of bounds
*** error within loop in function wes
> genr f[j+1]=(F[j] + m*T[j])*S[j+m-$pd]
I am getting the above error while trying to estimate the holt-winters from fcModels. Any
help?
Yes, I can see there's a bug in the wes() function whereby the
observation index can go out of bounds. That's fixed in a testing
version of the package, which you can install from the gretl console
via the command
pkg install https://users.wfu.edu/cottrell/gretl/fcModels.gfn
If that seems correct we can see about an "official" revision of fcModels.
Allin
3:30 a.m.
Hi,
How to get next 15 out of sample forecasts? I added time observations but not getting the
output.
12:44 p.m.
Am 03.03.2026 um 17:59 schrieb Cottrell, Allin:
On Tue, Mar 3, 2026 at 2:45 AM <ramakrishnasalagrama(a)gmail.com>
wrote:
> I am getting the above error while trying to estimate the holt-winters from fcModels.
Any help?
Yes, I can see there's a bug in the wes() function whereby the
observation index can go out of bounds. That's fixed in a testing
version of the package, which you can install from the gretl console
via the command
pkg install https://users.wfu.edu/cottrell/gretl/fcModels.gfn
If that seems correct we can see about an "official" revision of fcModels.
Hi Allin, the recent new email by Ramki S. brought me back to this
issue. The fcModels version on the package server is still 1.3 from
2023. What are your plans --if any-- for a revision? (Of course noting
that the actual package author is Yi-Nung Yang, who doesn't appear to be
maintaining this package anymore.)
cheers
sven
3:01 p.m.
Hi,
By running pkg install https://users.wfu.edu/cottrell/gretl/fcModels.gfn, there is no
error. However, this code is not populated in the server it seems. Further, this code also
not generating forecasts. Any fix? To teach time series models, this fcmodels is very
important for me.
4:44 p.m.
Hi,
I fixed the issue of wes in fcmodels. Anyone can test the code and confirm? New file....
https://drive.google.com/uc?export=download&id=1Iil_UEbRq6DQovYpwfZpn...
4:54 p.m.
OK, sorry, we'll make an effort to get a fixed fcModels into place.
Please try again tomorrow.
Allin
On Wed, Sep 2, 2026 at 11:01 AM Ramki S <ramakrishnasalagrama(a)gmail.com> wrote:
>
> Hi,
>
> By running pkg install https://users.wfu.edu/cottrell/gretl/fcModels.gfn, there is no
error. However, this code is not populated in the server it seems. Further, this code also
not generating forecasts. Any fix? To teach time series models, this fcmodels is very
important for me.
> _______________________________________________
> 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/
2:27 p.m.
Am 02.09.2026 um 18:54 schrieb Cottrell, Allin:
OK, sorry, we'll make an effort to get a fixed fcModels into
place.
Please try again tomorrow.
Hi Allin, first let me stress I'm not considering you as being
responsible for fixing or maintaining the fcModels package. But this
thread seems to be unfinished. The version on the package server is
still unchanged AFAICS. Are you planning to put your interim version there?
cheers
sven
4:49 p.m.
On Sun, Sep 20, 2026 at 10:28 AM Sven Schreiber
<sven.schreiber(a)fu-berlin.de> wrote:
Am 02.09.2026 um 18:54 schrieb Cottrell, Allin:
> OK, sorry, we'll make an effort to get a fixed fcModels into place.
> Please try again tomorrow.
>
Hi Allin, first let me stress I'm not considering you as being
responsible for fixing or maintaining the fcModels package. But this
thread seems to be unfinished. The version on the package server is
still unchanged AFAICS. Are you planning to put your interim version there?
Sorry, it's a bit complicated. I'll be happy to talk about this the
next chance we get.
Allin
11:33 a.m.
Am 21.09.2026 um 18:49 schrieb Cottrell, Allin:
On Sun, Sep 20, 2026 at 10:28 AM Sven Schreiber
<sven.schreiber(a)fu-berlin.de> wrote:
> Hi Allin, first let me stress I'm not considering you as being
> responsible for fixing or maintaining the fcModels package. But this
> thread seems to be unfinished. The version on the package server is
> still unchanged AFAICS. Are you planning to put your interim version there?
Sorry, it's a bit complicated. I'll be happy to talk about this the
next chance we get.
For the record for people out there who want to do double or triple
exponential smoothing aka Holt or Winters, with data-determined
smoothing parameters.
Thanks to Allin there are new versions of the latter packages on the
server, namely Holt.zip v2.0 and Winters.zip v2.2. Both provide fast and
continuous algorithms to estimate the optimal parameters (no grid
search). The included documentation is quite extensive, but judge for
yourself.
There are currently no plans anymore to patch the fcModels contributed
package in that area without the help of the actual package author.
cheers
sven
14
days inactive
233
days old
11 comments
4 participants
participants (4)
-
Cottrell, Allin -
ramakrishnasalagrama＠gmail.com -
Ramki S
-
Sven Schreiber