Open davidackerman opened 5 months ago
Hello,
I have noticed that the endianness of the labels array affects the ids that the mesher reports. For example
import zmesh import numpy as np for endian in ['<','>']: labels = np.zeros((11, 17, 19), dtype=np.dtype(f'{endian}i8')) labels[1:-1, 1:-1, 1:-1] = 1 mesher = zmesh.Mesher((4, 4, 40)) mesher.mesh(labels) print(f"{endian} endian, ids: {mesher.ids()}")
results in:
< endian, ids: [1] > endian, ids: [72057594037927936]
I can fix the problem by swapping the byteorder, but figured I'd point it out. I had noticed a similar fix in the skeletonization code.
Hello,
I have noticed that the endianness of the labels array affects the ids that the mesher reports. For example
results in:
I can fix the problem by swapping the byteorder, but figured I'd point it out. I had noticed a similar fix in the skeletonization code.