Error message in filter

Topics: Kalman Filtering
Jul 10, 2014 at 9:06 AM
Hi,

I'm started to use Ministry of Finance of Finland macro model with iris version 20140612 with Matlab 2014a and I get some error messages my colleagues using older versions do not.

In filter with option 'deviation', true I'll get this message:

_Subscript indices must either be real positive integers or
logicals.
Error in model/mykalman>xxSmoothMean (line 875)
[y2,f2,b2,e2,r] =
kalman.onestepbackmean(S,t,S.pe(:,1,t,1),S.a0(:,1,t,1),
...
Error in model/mykalman (line 338)
    s = xxSmoothMean(s);
Error in model/filter (line 235)
[obj,regOutp,hData] = mykalman(This,Inp,hData,likOpt);_

I did find out that S.d is empty with option 'true', but not with 'false'

I'd maneged to pass the error by doing the following in classlib\@model\mykalman.m
    [y2,f2,b2,e2,r] = kalman.onestepbackmean(S,t,S.pe(:,1,t,1),S.a0(:,1,t,1), ...
    S.f0(:,1,t,1),S.ydelta(:,1,t),S.d(:),r); %JR 3.7.2014
    %S.f0(:,1,t,1),S.ydelta(:,1,t),S.d(:,min(t,end)),r);

It seems to work with both true and false options, but should the S.d be empty at this point and am I causing some other problems.

Cheers,

Jukka
Coordinator
Jul 10, 2014 at 9:24 AM
Thanks, Jukka. This is a typo/bug. Can you please create an issue for this? I'll fix it.

Best,
Jaromir
Marked as answer by jaromirbenes on 7/10/2014 at 2:24 AM