Description of Kalm function

[charles marchetti]

Good morning,
I wanted to use the kalm function developed in scilab.
I used a model very simple in dimension 2.
However, by typing equations of the kalman filter, and
by making turned the algorithm on an iteration, I didn't get
the same results.
Then, I went on the site of Del Moral and I saw that
initial conditions must be used to initialize  updated
equation and not that of prediction
On a long interval of time, if the filter converges, it's not bad. Yet, with treatment on the estimation at the begining, it's obviously awfull. 
There are two schools: So the description of this function is not so clear... 

> > 
> > 
> > 
> >       
> 
> 
> -- 
> Sylvestre Ledru
> -------------------------
> The Scilab Consortium
> Digiteo
> Domaine de Voluceau
> Rocquencourt - B.P. 105
> 78153 Le Chesnay Cedex
> France
> 
> 
> 
>