indexvalue: Index–value representation of arrays
In AkselA/R-ymse: Ymse (Various)

Description Usage Arguments Details See Also Examples

Represent an array as columns of dimensional indices and value

1	indexvalue(x, reverse = FALSE)

`x`	an array or something that can be coerced into an array
`reverse`	logical; convert from Index–value representation to regular array representation?

An n-dimensional array will be unfolded to a n+1-column data.frame where the first n columns represent the indices of the n dimensions, and the last column gives the value found at each index tuple. The reverse process can also be performed.

latin_sq

arr <- array(1:(2*3*4), dim=c(2, 3, 4))
arr.is <- indexvalue(arr)

# can be used to permutate an array
indexvalue(arr.is[,c(2, 1, 3, 4)], rev=TRUE)
aperm(arr, c(2, 1, 3))

# can interpret values (symbols) as dimensional indices and permute them as well
arr2 <- array(rep(1:6, 4), dim=c(2, 3, 4))
arr2.is <- indexvalue(arr2)
indexvalue(arr2.is[,c(1, 2, 4, 3)], rev=TRUE)

# a latin square will produce an "orthogonal array"
set.seed(1)
lsq <- latin_sq(5)
iv <- indexvalue(lsq)
iv

# any permutation of a latin square is also a latin square
indexvalue(iv[, c(1, 3, 2)], reverse=TRUE)