A simple KD-Tree for numba using a ctypes wrapper around the scipy ckdtree
implementation.
The KD-Tree is usable in both python and numba nopython functions.
Once the query functions are compiled by numba, the implementation is just as fast as the original scipy version.
Note: Currently only a basic subset of the original ckdtree
interface is implemented.
pip install numba-kdtree
git clone https://github.com/mortacious/numba-kdtree.git
cd numba-kdtree
python setup.py install
import numpy as np
from numba_kdtree import KDTree
data = np.random.random(3_000_000).reshape(-1, 3)
kdtree = KDTree(data, leafsize=10)
# query the nearest neighbors of the first 100 points
distances, indices = kdtree.query(data[:100], k=30)
# query all points in a radius around the first 100 points
indices = kdtree.query_radius(data[:100], r=0.5, return_sorted=True)
The KDTree
can also be used from within numba functions
import numpy as np
from numba import njit
from numba_kdtree import KDTree
def numba_function_with_kdtree(kdtree, data):
for i in range(data.shape[0]):
distances, indices = kdtree.query(data[0], k=30)
#<Use the computed neighbors
data = np.random.random(3_000_000).reshape(-1, 3)
kdtree = KDTree(data, leafsize=10)
numba_function_with_kdtree(kdtree, data[:10000])
- Implement all scipy
ckdtree
functions - Fix the parallel query functions