[R-sig-Geo] masking z values (extract or subset on a 3-d array)

[Andy Bunn]

> > This is driving me nuts. I have a 3-D array that holds x,y,z 
> > coordinates from multiple experiments. I need to clean the 
> data before 
> > making it sp objects. I want to set all z-values < 0 to 0. In a 2-D 
> > matrix this is trivial e.g., x[,x[1,]<0] <- 0. But I'm 
> stumped on how 
> > to do this on a 3-D array without a loop. Given an array:
> > 
> >   foo <- array(rnorm(24), dim=c(4,3,2))
> >   dimnames(foo) <- 
> list(NULL,c('x','y','z'),c('Experiment1','Experiment2'))
> >   foo[,'z',]
> > 
> > How can I set foo[,'z',]<0 to 0?
> 
> a <- array(rnorm(64), c(4,4,4))
> inds <- which(a < 0, arr.ind=TRUE)
> a[inds] <- 0
> 

Thanks, but I'm looking to set only the values from one dimension to zero.
As above, I just want to set the 'z' values to 0 if they are less than one.
This works:

  foo <- array(rnorm(24), dim=c(4,3,2))
  dimnames(foo) <-list(NULL,c('x','y','z'),c('Experiment1','Experiment2'))
  foo[,'z',]
  #How can I set foo[,'z',]<0 to 0
  for(i in 1:2){
    foo[foo[,'z',i]<0,'z',i] = 0
  }
  foo

But it seems like one should be able to do this with apply or just '['. I'm
at a loss.

-A