Dear All,
I have a problem with running a VAR: i selected three 
endogenous variables and one dummy as exogenous, 8 lags chosen by lag 
selection option, with constant, trend and robust error options 
selected, but gretl is saying to me matrix is no positive definite: how 
can I sort this issue out?

Thanks

Regards

Mirko  

----Messaggio 
originale----
Da: paolo.chirico(a)unito.it
Data: 30/05/2012 10.57
A: 
"Gretl list"<gretl-users(a)lists.wfu.edu>
Ogg: Re: [Gretl-users] problem 
with kalman filter

Il 29/05/2012 18.39, Allin Cottrell ha scritto:
> 
On Tue, 29 May 2012, Paolo Chirico wrote:
>
>> I have run the following 
script for local linear model with quarterly
>> seasonality:
> [snip]
>

> you define a matrix Q:
>
> matrix Q = {s1^2, 0, 0, 0, 0; \

>               0, s2^2, 0, 0, 0; \
>               0, 0, s3^2, 0, 0; \

>               0, 0, 0, 0, 0; \
>               0, 0, 0, 0, 0}
>
> 
Then you do:
>
>> kalman
>>     obsy y
>>     obsymat z
>>     statemat 
T
>>     obsvar r
>>     statevar Q
>> end kalman --diffuse
>>
>> mle 
ll = ERR ? NA : $kalman_llt
>>     ERR = kfilter()
>>     params r s1 
s2 s3
>> end mle
> mle is spinning its wheels: you haven't provided any 
means of
> updating your Q matrix based on the current values of s1, s2 
and s3.
> You need to insert, just before the call to kfilter()
>

>        Q[1,1] = s1^2
>        Q[2,2] = s2^2
>        Q[3,3] = s3^2
>

> (There may be other problems but that's the most obvious one.)
>
> 
Allin Cottrell
> _______________________________________________
> 
Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.
edu/mailman/listinfo/gretl-users
>
>
You are right.
I have corrected 
the script and now it' ok.

Thanks

Paolo

-- 
Paolo Chirico

Ricercatore e Professore Aggregato
di Statistica Economica
Università 
di Torino
Via Maria Vittoria, 38
10123 Torino (Italy)



_______________________________________________
Gretl-users mailing 
list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.
edu/mailman/listinfo/gretl-users





Invita i tuoi amici e Tiscali ti premia! Il consiglio di un amico vale più di 
uno spot in TV. Per ogni nuovo abbonato 30 € di premio per te e per lui! Un 
amico al mese e parli e navighi sempre gratis: http://freelosophy.tiscali.it/

Reply via email to