[R-sig-Geo] Defining a grid for interpolations ?

[Mauricio Zambrano]

2008/5/27 Edzer Pebesma <edzer.pebesma at uni-muenster.de>:
> Mauricio Zambrano wrote:
>>
>> Dear Paul and Edzer,
>>
>> Thanks a lot for your answers.
>>
>> Before reading the solution proposed by Paul, I had tried with:
>>
>> # reading the boundary of the catchment
>> library(maptools)
>> catchment <- readShapePoly("only_catchment.shp")
>> catchment.grid <- spsample(catchment, cellsize=100, offset = c(0.5,
>> 0.5)) # reading the boundary of the catchment
>>
>>  and then I did an IDW interpolation with:
>>
>> pp.idw <- krige(PP_DAILY_MEAN_MM~1, meteo, catchment.grid)
>>
>> and it works.
>>
>> However, after reading your solution, I realised that I didn't use the
>> command:
>>
>> gridded(catchment.grid)
>>
>> but I got an interpolation anyway.
>>
>> Where is performed the interpolation when I don't use a gridded grid (
>> gridded(catchment.grid) ) ?
>>
>
> it is actually
>
> gridded(catchment.grid) <- TRUE

Sorry for the mistake.

>
> that changes the object, if it not gridded.
>
> For interpolation, it doesn't matter whether you loop over a set of points,
> or a set of points on a grid.

But if the object (catchment.grid) is not gridded, what points are
used for carrying out the interpolation "pp.idw <-
krige(PP_DAILY_MEAN_MM~1, meteo, catchment.grid "?

> --
> Edzer
>

Mauricio

Linux user  #454569 -- Ubuntu user #17469