How does IRIS handles missing observation in a measurement variable?

[Jesús Morales]

Hi! I’m using IRIS to estimate a model with some missing observation in one measurement variable. It works find, but I wonder how it handles that problem. Does it interpolate the missing date or does it uses the kalman filter to estimate the missing data?

Can IRIS be used to estimate a model with date of different frequencies (quarterly, yearly)?

[Jaromir Benes]

Hi Jesus,

You can think of it this way: Whenever there is a missing observation, the Kalman filter behaves as if the respective measurement equation were not in the model in that particular period. No information is used from that variable. And the observation for the missing variable is actually fitted by the Kalman filter based on all other observations and the model structure.

As for the other question: yes, you can design model where the measurement variables would represent different frequencies.

Hope this helps clarify the questions.

Best,

Jaromir

[Antti Ripatti]

So, you have time-varying coefficient matrix in the measurement equation?

--
You received this message because you are subscribed to the Google Groups "iris-toolbox-discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iris-toolbox...@googlegroups.com.
To post to this group, send email to iris-t...@googlegroups.com.
Visit this group at http://groups.google.com/group/iris-toolbox.
For more options, visit https://groups.google.com/groups/opt_out.

[Jaromir Benes]

Yes, exactly..

[Antti Ripatti]

Great! Hence, this would allow also mixed frequency observables?

[Jaromir Benes]

Yep..

[Jesús Morales]

·         Thank you very much! Can you give me a reference to learn a bit more about the technique?

[Jaromir Benes]

Do you mean how to use it in IRIS, how the Kalman filter theory?

--

[antti....@gmail.com]

I recommend Andrew Harvey's book on Structural Tine Series Models. It is much better tha Hamilton and has easy nitation.

--

[Серёжа Плотников]

Agree with Antti, here is some preview of that book

http://books.google.com/books?id=Kc6tnRHBwLcC&lpg=PP1&pg=PA100&redir_esc=y#v=onepage&q&f=false

Be happy,

Sergey

2013/12/20 <antti....@gmail.com>

[Jesús Morales]

Thank you very much! I have Harvey's book and I'm checking it.

El jueves, 19 de diciembre de 2013 16:02:50 UTC-4:30, Jesús Morales escribió: