Page 1 of 1

PDL's in State Space Models

Posted: Thu Jun 18, 2009 9:17 am
by dhanssen
Hello!

I am trying to use the sspace object, and I see in the code examples that usually the state/signal models are specified like the following:

log(passenger) = c(1) + sv1 + c(4)*sv2

@state sv1 = c(2)*sv1(-1) + c(3)*sv2(-1) + [var=exp(c(5))]

@state sv2 = sv1(-1)

However, in my model both the signal and state equation use pdl's:

@state sv1 = c(1)*sv1(-1) + PDL(LOG(x(-7)+1),6,2) PDL(y(-1)*100+1),4,2) PDL(z(-0)+1),4,2)

Is there a way to specify this in the state/space model framework?

Thank you so much for your help!

Re: PDL's in State Space Models

Posted: Thu Jun 18, 2009 9:24 am
by startz
Since a pdl is just a particular set of linear restrictions on lag coefficients, you could write out the model explicitly. Admittedly, it might be pretty time consuming.