2D Array Mirror

This is a sandbox/challenge project in preparation for the Camera Mirror project where this is used to create symetrical images from a live webcam feed.

The challenge

All of the data points on both the left and right 'sides' of the data when considered in a grid described by the ratio value need to be extracted, combined with it's own mirror annd returned as a new array.

Notes

• The output should cater for grids with odd and even number of columns.
• The central node should not be repeated.
• Do not assume each entry in the array will be unique.

Example data set

const source = [
  0, 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,
]

// For this example we will assume this ratio:
const ratio = [10, 4]

Expected output

const leftSideMirror = [
  4,  3,  2,  1,  0,  1,  2,  3,  4,
  14, 13, 12, 11, 10, 11, 12, 13, 14,
  24, 23, 22, 21, 20, 21, 22, 23, 24,
  34, 33, 32, 31, 30, 31, 32, 33, 34
]

const rightSideMirror = [
   9,  8,  7,  6,  5,  6,  7,  8,  9,
  19, 18, 17, 16, 15, 16, 17, 18, 19,
  29, 28, 27, 26, 25, 26, 27, 28, 29,
  39, 38, 37, 36, 35, 36, 37, 38, 39
]

Solution notes

Approach

• Determine the 'side' of the grid each node is on
• Filter each node into new left and right arrays
• Split new arrays into rows for that side
• Duplicate and reverse each row and combine with it's orignal
• Merge into single array output

How to calcualte 'position' in array matrix

To get column index for 0 based range

(i + 1) % ColumnCount

(i + 1) % ratio[0]

To get the row index, divide the index by the column count and round down

Math.floor(i / ratio[0])

matrix coords = [(i + 1) % ratio[0], Math.floor(i / ratio[0])]

Mirroring data

In order to get a consistant mirror from the vertical centeral axis of the matrix

Left side; combines the original followed by the mirror

Right side; combines the mirrored with the original

Odd and even column lengths

If column count is even

left side is full half (1-5)

right side is non inclusive of the left (6-10)

Each side is the same length

Odd column counts

In order to get the central pixel, both sides must include the central node

For left side; Divide row length and round up Math.ceil(ratio[0] / 2)

For right side; Divide row length and round down Math.floor(ratio[0] / 2)

Preventing central pixel duplicates

One of the row sets need to exclude the node in the center of the grid

For consistancy, the original data set is preserveed; the mirrored will exclude the center

This way, all manipulation is done on the same array