[R] Find mean of values in three-dimensional array

[Dénes Tóth]

On 06/15/2016 09:05 PM, peter dalgaard wrote:
 >
 >> On 15 Jun 2016, at 19:37 , Nick Tulli <nick.tulli.95 at gmail.com> wrote:
 >>
 >> Hey R-Help,
 >>
 >> I've got a three dimensional array which I pulled from a netcdf file.
 >> The data in array are the humidity values of locations in the United
 >> States over a time period. The three dimensions are [longitude,
 >> latitude, days], 141x81x92. My goal is to find the mean value at each
 >> longitude/latitude over the 92 day period.
 >>
 >> I could probably accomplish my goal by running a loop, but I'm sure
 >> that there is a much easier and more efficient way to accomplish the
 >> goal in R. Any suggestions?
 >
 > Dunno about fast, but the canonical way is apply(A, c(1,2), mean)

For "mean" and "sum", row/colMeans() is pretty fast and efficient. Note 
the 'dims' argument; you might also consider the aperm() function before 
the aggregation.

E.g.:

# create an array
x <- provideDimnames(array(rnorm(141*81*92), c(141, 81, 92)))
names(dimnames(x)) <- c("long", "lat", "days")

# collapse over days
str(rowMeans(x, dims = 2))

# collapse over lat
x_new <- aperm(x, c("lat", "long", "days"))
str(colMeans(x_new))

Cheers,
Denes


 >
 > E.g.
 >
 > (A <- array(1:24,c(2,3,4)))
 > apply(A, c(1,2), mean)
 > apply(A, c(1,3), mean)
 >
 > -pd
 >
 >>
 >>
 >> Thanks guys.
 >>
 >> ______________________________________________
 >> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
 >> https://stat.ethz.ch/mailman/listinfo/r-help
 >> PLEASE do read the posting guide 
http://www.R-project.org/posting-guide.html
 >> and provide commented, minimal, self-contained, reproducible code.
 >