12:23 p.m.
Hello
I've been using the Holt-Winters package but I cannot replicate my
Excel-calculation results with this package.
The starting value from the package for the trend is 245 mine is 166.396. By
running a regression on the full sample I get completely different results
for these starting values, although the same methodology as described in the
help description has been used.
For replication purposes I send my excel spreadsheet and I would be glad to
understand how these starting values have been effectively calculated.
Thank you for your help
Raul Gimeno
**
2:56 p.m.
I have not much time for testing today, but it seems that a correction I
included to avoid initial conditions very different from the first
observations of the series is not working well for your data. If you
change in the HoltWinters package the line
series yh1= (0.85*y<= $yhat && $yhat <=1.15*y) ? $yhat : y
simply to:
series yh1= $yhat
you will have results more similar (but not exactly) to yours (I think
we are using different initial conditions).
El 28/09/15 a las 14:23, Raul Gimeno escribió:
Hello
I've been using the Holt-Winters package but I cannot replicate my
Excel-calculation results with this package.
The starting value from the package for the trend is 245 mine is 166.396. By
running a regression on the full sample I get completely different results
for these starting values, although the same methodology as described in the
help description has been used.
For replication purposes I send my excel spreadsheet and I would be glad to
understand how these starting values have been effectively calculated.
Thank you for your help
Raul Gimeno
**
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
--
Ignacio Díaz-Emparanza
Departamento de Economía Aplicada III (Econometría y Estadística)
Universidad del País Vasco - Euskalherriko Unibertsitatea, UPV/EHU
Tfno: (+34) 94 601 3732
http://www.ehu.eus/ea3
8:40 p.m.
Some time ago I came across a similar problem when using this function
for a class. At the time I made a note to myself saying that the
problem was in the definition of lobs, which I changed to:
scalar lobs=lastobs(y)
I hope this helps
On 28 September 2015 at 15:56, Ignacio Diaz-Emparanza
<ignacio.diaz-emparanza(a)ehu.eus> wrote:
I have not much time for testing today, but it seems that a
correction I
included to avoid initial conditions very different from the first
observations of the series is not working well for your data. If you change
in the HoltWinters package the line
series yh1= (0.85*y<= $yhat && $yhat <=1.15*y) ? $yhat : y
simply to:
series yh1= $yhat
you will have results more similar (but not exactly) to yours (I think we
are using different initial conditions).
El 28/09/15 a las 14:23, Raul Gimeno escribió:
>
> Hello
>
> I've been using the Holt-Winters package but I cannot replicate my
> Excel-calculation results with this package.
> The starting value from the package for the trend is 245 mine is 166.396.
> By
> running a regression on the full sample I get completely different results
> for these starting values, although the same methodology as described in
> the
> help description has been used.
> For replication purposes I send my excel spreadsheet and I would be glad
> to
> understand how these starting values have been effectively calculated.
> Thank you for your help
> Raul Gimeno
>
> **
>
>
> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
--
Ignacio Díaz-Emparanza
Departamento de Economía Aplicada III (Econometría y Estadística)
Universidad del País Vasco - Euskalherriko Unibertsitatea, UPV/EHU
Tfno: (+34) 94 601 3732
http://www.ehu.eus/ea3
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
10:33 a.m.
El 28/09/15 a las 22:40, Pedro Bação escribió:
Some time ago I came across a similar problem when using this
function
for a class. At the time I made a note to myself saying that the
problem was in the definition of lobs, which I changed to:
scalar lobs=lastobs(y)
I hope this helps
Yes, I detected this problem some months ago and included some changes
for treating with missing observations as well.
With respect to Raul problem, apart from the change in line
series yh1= (0.85*y<= $yhat && $yhat <=1.15*y) ? $yhat : y
to:
series yh1= $yhat
which I commented yesterday, I see that the differences in calculations
between Raul's excel functions and this package was because of the
different initial observations. As recently reported for the 'movavg'
command (exponential moving average) I was also ignoring the first
observation of the series. I have corrected this and committed the
change (It is in the staging area until Allin aproval:
http://ricardo.ecn.wfu.edu/gretl/staging_fnfiles/).Now the results are
the same.
I also think we need more flexibility for the initial conditions, I will
work on this.
--
Ignacio Díaz-Emparanza
Departamento de Economía Aplicada III (Econometría y Estadística)
Universidad del País Vasco - Euskalherriko Unibertsitatea, UPV/EHU
Tfno: (+34) 94 601 3732
http://www.ehu.eus/ea3
1:48 p.m.
Dear Ignacio,
I'm trying to estimate the parameters of a HW method by means of MLE
command:
mle ll = -1/2*log(2*$pi*s2)-1/2*(e^2)/s2
loop i=1+s...$nobs
(HW method)
endloop
params alpha beta gamma
end mle
Unfortunately the loop command is not permitted by mle commnd!
Have you any idea/advice to solve this problem?
Best regards,
Paolo
--
Paolo Chirico
Assistant Professor of Economic Statistics
Dept. of Economics and Statistics "Cognetti de Martiis"
Campus Luigi Einaudi, University of Turin
ITALY
2895
days inactive
3134
days old
4 comments
4 participants
participants (4)
-
Ignacio Diaz-Emparanza -
Paolo Chirico -
Pedro Bação -
Raul Gimeno