Thanks Riccardo for the quick reply and your suggestion.

I'm specifically want to use ARMA, because I use it in my thesis and 
benchmark it against a bunch of other interpolators I've implemented. 
Once I get that working I'll definitely look into your Kalman-filter 
suggestion.

Do you maybe know how to add the time-lags to the DATASET? The gretl API 
documentation (and the examples in the source code) are not very 
informative on this matter. And how do I set t1 and t2 of the DATSET so 
that the ARMA model and get_forecast knows where the samples start and 
end (since there are now 2 places where the data starts and ends, not 
just 1 as with the usual DATASET used to do a simple forecast).

Regards
Chris

On 2014/05/15 08:02 AM, Riccardo (Jack) Lucchetti wrote:
> On Thu, 15 May 2014, GOO Creations wrote:
>
>> Hi,
>>
>> I'm not sure if this is possible in gretl. I want to interpolate a gap
>> of samples with ARMA by using the value to the left and right of the
>> gap. If I have data like this:
>>
>> Time lag:    1    2    3    4    5    6    7    8    9    10 11
>> Values:        0    0.1    0.2    0.3    *    *    *    0.3 0.2    
>> 0.1    0
>>
>> The values marked with a star are the gap I want to interpolate (at lag
>> 5, 6 and 7: the values should be something like 0.4, 0.5, 0.4 after
>> interpolation). How would you go about creating a gretl DATASET with
>> these values? And will the get_forecast function work on "predicting"
>> the interpolated samples?
>> I've always used ARMA for out-of-sample forecasts, but never for
>> interpolation, so I'm not sure if this is possible.
>
> This is absolutely possible via an ARMA model. However, my advice 
> would be to use a state space model and the Kalman filter, which 
> provide a very flexible, natural and flexible framework for problems 
> of this type.
>
>
> -------------------------------------------------------
>   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)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-devel

Thanks Riccardo for the quick reply and your suggestion.

I'm specifically want to use ARMA, because I use it in my thesis and benchmark it against a bunch of other interpolators I've implemented. Once I get that working I'll definitely look into your Kalman-filter suggestion.

Do you maybe know how to add the time-lags to the DATASET? The gretl API documentation (and the examples in the source code) are not very informative on this matter. And how do I set t1 and t2 of the DATSET so that the ARMA model and get_forecast knows where the samples start and end (since there are now 2 places where the data starts and ends, not just 1 as with the usual DATASET used to do a simple forecast).

Regards
Chris

On 2014/05/15 08:02 AM, Riccardo (Jack) Lucchetti wrote:
On Thu, 15 May 2014, GOO Creations wrote:

Hi,

I'm not sure if this is possible in gretl. I want to interpolate a gap
of samples with ARMA by using the value to the left and right of the
gap. If I have data like this:

Time lag:    1    2    3    4    5    6    7    8    9    10    11
Values:        0    0.1    0.2    0.3    *    *    *    0.3 0.2    0.1    0

The values marked with a star are the gap I want to interpolate (at lag
5, 6 and 7: the values should be something like 0.4, 0.5, 0.4 after
interpolation). How would you go about creating a gretl DATASET with
these values? And will the get_forecast function work on "predicting"
the interpolated samples?
I've always used ARMA for out-of-sample forecasts, but never for
interpolation, so I'm not sure if this is possible.

This is absolutely possible via an ARMA model. However, my advice would be to use a state space model and the Kalman filter, which provide a very flexible, natural and flexible framework for problems of this type.


-------------------------------------------------------
  Riccardo (Jack) Lucchetti
  Dipartimento di Scienze Economiche e Sociali (DiSES)

  Università Politecnica delle Marche
  (formerly known as Università di Ancona)

  [email protected]
  http://www2.econ.univpm.it/servizi/hpp/lucchetti
-------------------------------------------------------


_______________________________________________
Gretl-devel mailing list
[email protected]
http://lists.wfu.edu/mailman/listinfo/gretl-devel

Reply via email to