neatnearestneighbour 1.0.0

A nearest neighbour algorithm written in pure Python.

A Nearest Neighbour algorithm written in pure python. The intention is to use this package with PyPy. The hot loop in the algorithm are perfect for PyPy's just-in-time compilation and results in fast performance.

To run create an instance of NearestNeighbour and call run(). The output is a generator of items in order by finding the next nearest neighbour of each item. The starting item is randomized.

The input items can be a list of anything, because the distance function also has to be provided. The distance function must be able to calculate the distance between any items in the input list.

An example:

from neatnearestneighbour import NearestNeighbour
items = [10, 20, 15, 5, 100, 2, 30, 50]
distance_fn = lambda a, b: abs(b - a)

nn = NearestNeighbour(
    items,
    distance_fn,
)
neighbours = nn.run()
print(list(neighbours))
# [50, 30, 20, 15, 10, 5, 2, 100]