R help - Nov 2008 - Fill vector filed NAs data

Good afternoon,
                             I am dealing with a 2d vector field on a
regular grid. The field has sparse NA values
(t near the domain borders and not isolated),
which I am trying to substitute with
interpolated values of the nearest point.
What I am looking for is to replace the
values with a weighted mean, where the
weights are function of the distance from the
regular values.
I wonder if R as an implemented function for this I have looked the zoo,
the akima and the fileds packages but I have not found it.
Thanks in advance,
                                   Luca


*********************************************************
dr. Luca Mortarini               l.mortarini at isac.cnr.it
Istituto di Scienze dell?Atmosfera e del Clima ? ISAC
Consiglio Nazionale delle Ricerche ? C.N.R.
&
Universit? del Piemonte Orientale
Dipartimento di Scienze e Tecnologie Avanzate

On Nov 4, 2008, at 8:02 AM, Luca Mortarini wrote:
> I am dealing with a 2d vector field on a
> regular grid. The field has sparse NA values
> (t near the domain borders and not isolated),
> which I am trying to substitute with
> interpolated values of the nearest point.
> What I am looking for is to replace the
> values with a weighted mean, where the
> weights are function of the distance from the
> regular values.
> I wonder if R as an implemented function for this I have looked the  
> zoo,
> the akima and the fileds packages but I have not found it.
> Thanks in advance,
>                                   Luca
I have not figured out what you mean by "substitut[ing] with
interpolated values of the nearest point". Would have seemed that  
interpolation would require at a minimum two interior points. If these  
are "not isolated" and at or near borders, it would also appear that  
extrapolation might be a more appropriate term.

Quibbles aside .... Have you looked at the locfit package? Regarding  
the vector-valued aspect of this field, perhaps you could approach it  
by analyzing the projections on single dimensions at a time.

-- 
David Winsemius
Heritage Labs