Gavin Wiggins

AboutNotesPhotos


Flat indexing an N-dimensional array

Written on August 7, 2024

An N-dimensional or multi-dimensional array is a data structure with one or more dimensions. An index for each dimension must be provided to access a particular element's value. Since the array's data is stored in one-dimensional memory, the indices must be converted to a flat index (one-dimension) to access the corresponding memory location.

The following function calculates the flat index of an element in an N-dimensional array. It is applicable for row-major order and zero-based indexing. The indices input parameter is the index for each dimension and the shape input parameter is the size of the corresponding dimensions. For example, a 2D array may have indices: [1, 3] which is the element in the second row and fourth column. Also, a 2D array with shape: [4, 8] describes the array as having four rows and eight columns.

func flatIndex(indices: [Int], shape: [Int]) -> Int {
    var index = 0
    var stride = 1

    for i in (0..<shape.count).reversed() {
        index += indices[i] * stride
        stride *= shape[i]
    }

    return index
}

The function is based on the formula shown below where indices are \(i_1, i_2, ..., i_N\) and dimensions are \(d_1, d_2, ..., d_N\).

$$ \text{flat index} = i_1 + (i_2 \times d_1) + (i_3 \times d_1 \times d_2) + ... + (i_N \times d_1 \times d_2 \times ... \times d_{N-1}) $$

The example below calculates the flat index for a 2D array.

let a = [1, 2, 3, 4,
         5, 6, 7, 8]

let idx = flatIndex(indices: [1, 2], shape: [2, 4])

print("Flat index is \(idx) which is value \(a[idx]) in the 2D array")
// This prints:
// Flat index is 6 which is value 7 in the 2D array

This example calculates the flat index for a 3D array.

let a = [1, 2, 3, 4,
         5, 6, 7, 8,

         9, 10, 11, 12,
         13, 14, 15, 16]

let idx = flatIndex(indices: [1, 0, 2], shape: [2, 2, 4])

print("Flat index is \(idx) which is value \(a[idx]) in the 3D array\n")
// This prints:
// Flat index is 10 which is value 11 in the 3D array

This example is for a 4D array.

let a = [1, 2, 3, 4,
         5, 6, 7, 8,

         9, 10, 11, 12,
         13, 14, 15, 16,

         17, 18, 19, 20,
         21, 22, 23, 24,


         25, 26, 27, 28,
         29, 30, 31, 32,

         33, 34, 35, 36,
         37, 38, 39, 40,

         41, 42, 43, 44,
         45, 46, 47, 48]

let idx = flatIndex(indices: [1, 2, 1, 3], shape: [2, 3, 2, 4])

print("Flat index is \(idx) which is value \(a[idx]) in the 4D array\n")
// This prints:
// Flat index is 47 which is value 48 in the 4D array

The flatIndex function also works with a 1D array as shown here.

let a = [1, 2, 3, 4, 5]
let idx = flatIndex(indices: [1], shape: [a.count])

print("Flat index is \(idx) which is value \(a[idx]) in the 1D array\n")
// This prints:
// Flat index is 1 which is value 2 in the 1D array

Gavin Wiggins © 2024.
Made on a Mac with Genja. Hosted on GitHub Pages.