[R-sig-Geo] MODIS Whittaker Smoothing multicore support

[Amit Boshale]

Hi Forrest, that worked! Thank you so much!I was always getting an error that sendCall function was not found, I just had to load snow package and everything is fine. 
Thanks!Amit







     Forrest Stevens <forrest at ufl.edu> schrieb am 15:26 Dienstag, 31.März 2015:
   

 

Hi Amit, I can provide you with some help to get smoothing working to
leverage multiple cores.  This is already baked into the
whittaker.raster() function of our MODIS package and it makes sense to
use it since the problem is of the "embarrassingly parallel" variety.
To leverage multiple cores you can do something like the following:


#beginCluster(type="SOCK",exclude="MODIS")
##  Alternatively:
nodes <- 4
beginCluster(nodes)
system.time(whittaker.raster(vi=vi, wt=qa, inT=time,
timeInfo=timeInfo, groupYears=FALSE,
outDirPath=paste(root_path,"output/",sep="")))
endCluster()

##  Where vi, qa, time, timeInfo are all arguments you can learn about
in the whittaker.raster() documentation.


Hope this helps!

Sincerely,
Forrest


On Tue, Mar 31, 2015 at 5:33 AM Amit Boshale via R-sig-Geo
<r-sig-geo at r-project.org> wrote:
>
> Dear Matteo,
> Are there any plans to enable multicore in Whittaker smoothing function? I use your MODIS package on regular basis and multicore calculations would make my life easier.
>
> Keep up the great work
> Amit
>        [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo


  
	[[alternative HTML version deleted]]