[R-sig-eco] memory issues using Tps (package fields)

[Robert J. Hijmans]

Hi Jorge,

I might work if you specify "dem" as an independent variable rather
than as a co-variable (z). (If you really want it to be a co-variable
it would take a bit of tweaking of the interpolate function). You also
need to add "xyOnly = F" to predict (not obvious from docs).

data<-read.table("precp00-09ContSampleDem.csv", sep="," , header=TRUE)
fit<-Tps( data[ , c("x","y", "dem") ],  data[,"Ene"] )
elev=raster("dem_col.asc")
result<-interpolate(elev, fit, xyOnly=FALSE)

(I cannot test this as I do not have (a sample of) your data)

Robert

On Thu, Jun 17, 2010 at 8:33 PM, Jorge Velasquez
<jorge at life.bio.sunysb.edu> wrote:
> Hi Robert,
>
> I tried the solution you suggested but get the following error:
>
> Error in predict.Krig(model, blockvals, ...) :
>  Need to specifify drop.Z as TRUE or pass Z values
>
> This is the code I used:
>
> data<-read.table("precp00-09ContSampleDem.csv",sep=",",header=TRUE)
> fit<-Tps(data[,c("x","y")],data[,"Ene"],z=data$dem)
> elev=raster("dem_col.asc")
> result<-interpolate(elev,fit)
>
> Many thanks indeed for your reply,
>
> Jorge
>
> On Thu, Jun 17, 2010 at 3:49 PM, Robert J. Hijmans <r.hijmans at gmail.com> wrote:
>> Jorge,
>> You could use the function 'interpolate' in 'raster'. It makes
>> predictions for smaller chunks of data and puts these back together.
>> There is an example for Tps. You would need to provide elevation data
>> (as a RasterLayer) as the first argument.
>> Robert
>>
>> On Thu, Jun 17, 2010 at 1:25 PM, Jorge Velasquez
>> <jorge at life.bio.sunysb.edu> wrote:
>>> Hi there,
>>>
>>> I'm trying to run a interpolation of climate using the functionTps
>>> (Thin plated splines) from the package fields, but keep running into
>>> the following error:
>>>
>>> Error: cannot allocate vector of size 488 Kb
>>> In addition: Warning messages:
>>> 1: In names(y) <- make.unique(cols) :
>>>  Reached total allocation of 3583Mb: see help(memory.size)
>>> 2: In names(y) <- make.unique(cols) :
>>>  Reached total allocation of 3583Mb: see help(memory.size)
>>> 3: In names(y) <- make.unique(cols) :
>>>  Reached total allocation of 3583Mb: see help(memory.size)
>>> 4: In names(y) <- make.unique(cols) :
>>>  Reached total allocation of 3583Mb: see help(memory.size)
>>> Error in as.matrix(expand.grid(grid.list)) :
>>>  error in evaluating the argument 'x' in selecting a method for
>>> function 'as.matrix'
>>>
>>> The code I'm running is:
>>>
>>> predGrid<-read.table("demaspoint_id8.csv" ,sep="," ,header=TRUE)
>>> data<-read.table("precp00-09ContSampleDem.csv",sep="," ,header=TRUE)
>>> fit<-Tps(data[,c("x","y")],data[,"Ene"],Z=data$dem)
>>> result<-predict.surface(fit,grid.list=list(x=predGrid[,c("Longitude","Latitude")],Z=predGrid[,"GRID_CODE"]))
>>>
>>> The file predGrid has ~65000 rows. The error is persistent even using
>>> a machine with 8GB Ram, double xeon processor, and the 64bit version
>>> of R, although the error pasted above is for the 32 bit version.
>>>
>>> Is there any solution to this? Am I missing something?
>>>
>>> Thanks,
>>>
>>> Jorge
>>> --
>>> ---
>>> Jorge Velasquez
>>> Department of Ecology and Evolution
>>> Stony Brook University
>>> Stony Brook, NY 11974
>>> Phone:(631) 632-1445
>>> http://life.bio.sunysb.edu/~jorge/
>>>
>>> _______________________________________________
>>> R-sig-ecology mailing list
>>> R-sig-ecology at r-project.org
>>> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
>>>
>>
>>
>
>
>
> --
> ---
> Jorge Velasquez
> Department of Ecology and Evolution
> Stony Brook University
> Stony Brook, NY 11974
> Phone:(631) 632-1445
> http://life.bio.sunysb.edu/~jorge/
>