R/tsk.R

In ltsk: Local Time Space Kriging

tsk <-
function(query,obs,subset=T,nmin=3,nmax=20,xcoord='x',ycoord='y',tcoord='t',zcoord='z',
		vth=NULL,vlen=NULL,llim=c(3,3),verbose=T,Large=2000,future=T)
{
# seed <- round(runif(1) * 1000000)
 l.query <- check_input(query,xcoord,ycoord,tcoord,zcoord)
 l.query <- check_na(l.query[,c(xcoord,ycoord,tcoord)],'query')
 l.obs <- check_input(obs,xcoord,ycoord,tcoord,zcoord)
 l.obs <- check_na(l.obs,'observed')
 
# if(is.null(cl)){
#	set.seed(seed=seed)
	## Use single core mode
 	out <- working.tsk(q0=l.query,obs=l.obs,subset=subset,nmin=nmin,nmax=nmax,vth=vth,vlen=vlen,
		llim=llim,verbose=verbose,Large=Large,future=future)
# }
#  else if ("cluster" %in% class(cl)){
# 	## Use multiple core mode
#  	clusterSetRNGStream(cl,seed)
# 	pwd <- getwd()
#  	clusterCall(cl,setwd,dir=pwd)
# 	clusterEvalQ(cl,library(ltsk))
# 
# 	res <- partUtil(l.obs,l.query,length(cl),th,xcoord=xcoord,ycoord=ycoord,tcoord=tcoord)
# 	ll.query <- vector('list',length(res$query))
# 	ll.obs <- vector('list',length(res$obs))
# 	ll.order <- vector('list',length(res$query))
# 	toMat <- function(m,i)
# 	{
# 	  if(length(i)>1){
# 	    mo <- m[i,]
# 	  }
# 	  else{
# 	    mo <- matrix(m[i,],ncol=ncol(m))
# 	  }
# 	  mo
# 	}
# 	for(i in 1:length(ll.query)){
# 	  ll.query[[i]] <- toMat(l.query,res$query[[i]])
# 	  ll.obs[[i]] <- toMat(l.obs,res$obs[[i]])
# 	  ll.order[[i]] <- res$query[[i]]
# 	}
# 	
#   
# 	ll.args <- list(th=th,vth=vth,vlen=vlen,llim=llim,verbose=verbose,Large=Large,future=future)
# 	out1 <- clusterMap(cl=cl,fun=working.tsk,ll.query,ll.obs,MoreArgs=ll.args)
# 	r.order <- do.call(c,ll.order)
# 	out <- matrix(unlist(out1),ncol=3,byrow=T)
# 	out <- out[order(r.order),]
# 	
#  }
#  else{
#    stop(cl," is not a cluster\n")
#  }
 colnames(out$krig) <- c('krig','se','flag')
 out
}