neu_vol.backends.cloudvolume¶

Read precomputed volumes written by CloudVolume, whose chunks are .gz-suffixed.

CloudVolume gzips each chunk and appends .gz to the key, which the precomputed spec does not address: tensorstore requests the unsuffixed key and returns the fill value, so the volume reads as all zeros and nothing raises. detect.detect_backend returns PRECOMPUTED_GZ for such volumes, which routes here, to the library that wrote them.

cloud-volume is optional and not in the main environment: it pins DracoPy<2, which would downgrade the Draco codec neu-morpho’s meshes need. Use a separate environment holding blockrun and neu-vol but not neu-morpho. Read-only.

Attributes¶

TAG

Classes¶

CloudVolumeBackend

Read-only (z, y, x) view over a CloudVolume-written precomputed volume.

Module Contents¶

class neu_vol.backends.cloudvolume.CloudVolumeBackend(spec: Mapping[str, Any])¶

Read-only (z, y, x) view over a CloudVolume-written precomputed volume.

classmethod open(spec: Mapping[str, Any]) → CloudVolumeBackend¶

read_region(region: neu_vol.backends.base.Region) → numpy.ndarray¶

to_spec() → dict[str, Any]¶

write_region(region: neu_vol.backends.base.Region, data: numpy.ndarray) → None¶

property chunks: tuple[int, ...]¶

property dtype: numpy.dtype¶

property shape: tuple[int, ...]¶

neu_vol.backends.cloudvolume.TAG = 'neuroglancer_precomputed_gz'¶