warp.volume_voxel_count

warp.volume_voxel_count(id: uint64) → int32

Kernel

Python

Return the number of indexable voxels in the volume given by id.

For NanoVDB OnIndex and OnIndexMask grids, this is the active voxel count. For Index and IndexMask grids and for classical value grids, it is the number of allocated leaf nodes multiplied by 512, because every leaf slot is indexable. When the true count fits in int32, this is the required size of a per-voxel voxel_data array (as used by volume_sample_index()) and the exclusive upper bound of indices returned by volume_lookup_index().

The result is capped at 2**31 - 1. A larger volume cannot be fully represented by these int32 index APIs: lookup indices may overflow, and the capped count is not sufficient to allocate data for every underlying slot. Not differentiable.

Parameters:

id – The id of a warp.Volume.

Returns:

The number of indexable voxels, capped at 2**31 - 1.

Example

@wp.kernel
def count(vid: wp.uint64, out: wp.array[wp.int32]):
    out[0] = wp.volume_voxel_count(vid)

volume = wp.Volume.load_from_numpy(np.zeros((2, 2, 2), dtype=np.float32), voxel_size=1.0, bg_value=0.0)
out = wp.zeros(1, dtype=wp.int32)
wp.launch(count, dim=1, inputs=[volume.id], outputs=[out])
print(int(out.numpy()[0]))

512