The Hilbert module contains these sub-modules:
- normalize - To prepare data for transformation.
- permutation - To reorder point coordinates as a means of generating alternate Hilbert curves.
- point_list - Prepare points from i32 or f64 data, deriving and applying a consistent transform to each.
- point - Represents an N-dimensional point suitable to be transformed and sorted by the Hilbert Curve.
- transform - Performs the core Hilbert Curve transform and its inverse.
normalize module contains
These structs are used to normalize point coordinates to prepare them for the Hilbert transform.
permutation module contains the
Permutation struct, which is used to
reorder the coordinates in one or more
Points in a consistent way.
This permits the generation of multiple Hilbert Curves, each rendering
the points in a different order. With multiple curves, you can improve the
accuracy of K-Nearest Nighbor searches and clustering algorithms.
point module defines the
Point struct, which represents an N-dimensional point.
Points are suitable for the Hilbert Transformation, because they do not have negative values
and the total number of bits required to represent each coordinate is controlled by the construction process.
point_list module can convert point-like data in forms defined by the caller
Point structs. It defines the functions
They can detect the range of values in the input data and the number of bits required
to code those values without losing precision, then translate and scale those values
in the process of creating
Points created in this way are guaranteed
to be acceptable to the Hilbert Curve transformation.
transform module holds the functions which perform the forward and inverse Hilbert Transform.
A more convenient way to prepare for and perform the forward transform is to create