This package provides functions to convert between Morton number (a.k.a. Z-order), Cartesian coordinates, and quadtree and octree coordinates.
Say for example you have a 4x4 matrix. The sixteen cells could be addressed in each of the following three ways.
Morton order:
1 | 2 | 5 | 6 |
3 | 4 | 7 | 8 |
9 | 10 | 13 | 14 |
11 | 12 | 15 | 16 |
Cartesian coordinates:
1,1 | 2,1 | 3,1 | 4,1 |
1,2 | 2,2 | 3,2 | 4,2 |
1,3 | 2,3 | 3,3 | 4,3 |
1,4 | 2,4 | 3,4 | 4,4 |
Quadtree coordinates:
1,1 | 1,2 | 2,1 | 2,2 |
1,3 | 1,4 | 2,3 | 2,4 |
3,1 | 3,2 | 4,1 | 4,2 |
3,3 | 3,4 | 4,3 | 4,4 |
To convert from Morton to Cartesian, use the morton2cartesian
function:
julia> Pkg.add("Morton")
julia> using Morton
julia> morton2cartesian(13)
2-element Array{Int64,1}:
3
3
Similarly, one can convert from Morton to quadtree, or Cartesian to quadtree:
julia> morton2tree(13)
2-element Array{Int64,1}:
4
1
julia> cartesian2tree([3,3])
2-element Array{Int64,1}:
4
1
Of course each of the functions can be reversed:
julia> cartesian2morton([3,3])
13
julia> tree2morton([4,1])
13
julia> tree2cartesian([4,1])
2-element Array{Int64,1}:
3
3
Corresponding functions also exist for three dimensional matrices (i.e.
octrees). Simply replace the 2 with a 3: morton3cartesian
, morton3tree
,
etc.
There are also un-exported N-dimensional functions to convert between tree and
Morton, and tree and Cartesian (e.g. Morton._treeNmorton
). Please let me
know if you have a clever way to convert directly between Morton and Cartesian
in arbitrary dimensions!
Ben Arthur, [email protected]
Scientific Computing
Janelia Research Campus
Howard Hughes Medical Institute