crantpy.viz.l2 module

Visualization module for CRANTBpy.

crantpy.viz.l2.chunks_to_nm(xyz_ch, vol, voxel_resolution=None)

Map a chunk location to Euclidean space.

Parameters:

• xyz_ch (array-like) – (N, 3) array of chunk indices.

• vol (cloudvolume.CloudVolume) – CloudVolume object associated with the chunked space.

• voxel_resolution (list, optional) – Voxel resolution. If None, use SCALE_X, SCALE_Y, SCALE_Z.

Returns:

(N, 3) array of spatial points.

Return type:

np.array

crantpy.viz.l2.find_anchor_loc(neurons, dataset=None, max_threads=4, progress=True)

Find a representative coordinate for neuron(s) using the L2 cache.

This works by querying the L2 cache and using the representative coordinate for the largest L2 chunk for each neuron.

Parameters:

• neurons (int, str, list, np.ndarray, or NeuronCriteria) – Root ID(s) to get coordinate for. Can be a single root ID, a list of root IDs, or an instance of NeuronCriteria.

• dataset (str, optional) – Dataset to query. If None, will use the default dataset.

• max_threads (int) – Number of parallel threads to use for batch queries.

• progress (bool) – Whether to show a progress bar.

Returns:

DataFrame with columns: root_id, x, y, z.

Return type:

pandas.DataFrame

crantpy.viz.l2.get_l2_chunk_info(neurons, dataset=None, progress=True, chunk_size=2000)

Fetch info for L2 chunks associated with given neuron(s).

Parameters:

• neurons (int, str, list, np.ndarray, or NeuronCriteria) – Neurons to fetch L2 chunk info for. Can be a single root ID, a list of root IDs, or an instance of NeuronCriteria.

• dataset (str, optional) – Dataset to fetch info from.

• progress (bool) – Whether to show a progress bar.

• chunk_size (int) – Number of L2 IDs per query.

Returns:

DataFrame with L2 chunk info (coordinates, vectors, size).

Return type:

pandas.DataFrame

crantpy.viz.l2.get_l2_dotprops(root_ids, min_size=None, sample=False, omit_failures=None, progress=True, max_threads=10, dataset=None, **kwargs)

Generate dotprops from L2 chunks for given neuron(s).

Parameters:

• root_ids (int, str, list, np.ndarray, or NeuronCriteria) – Root ID(s) of the FlyWire neuron(s) to generate dotprops for.

• min_size (int, optional) – Minimum size (in nm^3) for the L2 chunks. Smaller chunks will be ignored.

• sample (float (0 < sample < 1), optional) – If float, will create Dotprops based on a fractional sample of the L2 chunks.

• omit_failures (bool, optional) – Behaviour when dotprops generation fails. None (default) raises, True skips, False returns empty Dotprops.

• progress (bool) – Whether to show a progress bar.

• max_threads (int) – Number of parallel requests to make when fetching the L2 IDs.

• dataset (str, optional) – Against which FlyWire dataset to query. If None, will use the default dataset.

• **kwargs – Additional keyword arguments passed to Dotprops initialization.

Returns:

List of Dotprops.

Return type:

navis.NeuronList

crantpy.viz.l2.get_l2_graph(root_ids, dataset=None, progress=True, max_threads=4)

Fetch L2 graph(s) for given neuron(s).

Parameters:

• root_ids (int, str, list, np.ndarray, or NeuronCriteria) – FlyWire root ID(s) for which to fetch the L2 graphs.

• dataset (str, optional) – Against which FlyWire dataset to query. If None, will use the default dataset.

• progress (bool) – Whether to show a progress bar.

• max_threads (int) – Number of parallel threads to use for batch queries.

Returns:

The L2 graph or list thereof.

Return type:

networkx.Graph or list of networkx.Graph

crantpy.viz.l2.get_l2_info(neurons, dataset=None, progress=True, max_threads=4)

Fetch basic info for given neuron(s) using the L2 cache.

Parameters:

• neurons (int, str, list or NeuronCriteria) – Neurons to fetch info for. Can be a single root ID, a list of root IDs, or an instance of NeuronCriteria.

• dataset (str, optional) – Dataset to fetch info from.

• progress (bool) – Whether to show a progress bar.

• max_threads (int) – Number of parallel requests to make.

Returns:

DataFrame with basic L2 information for the given neuron(s). - length_um is the underestimated sum of the max diameter across all L2 chunks - bounds_nm is the rough bounding box based on the representative coordinates of the L2 chunks - chunks_missing is the number of L2 chunks not present in the L2 cache.

Return type:

pandas.DataFrame

crantpy.viz.l2.get_l2_meshes(x, threads=10, progress=True, dataset=None)

Fetch L2 meshes for a single neuron or NeuronCriteria in CRANTb.

Parameters:

• x (int | str | NeuronCriteria) – Root ID or NeuronCriteria. Must not be a list.

• threads (int)

• progress (bool)

• dataset (str, optional) – Against which CRANTb dataset to query. If None, will use the default dataset.

Return type:

navis.NeuronList

crantpy.viz.l2.get_l2_skeleton(root_ids, refine=True, drop_missing=True, l2_node_ids=False, omit_failures=None, progress=True, max_threads=4, dataset=None, **kwargs)

Generate skeleton(s) from L2 graph(s) for given neuron(s).

Parameters:

• root_ids (int, str, list, np.ndarray, or NeuronCriteria) – Root ID(s) of the CRANTb neuron(s) to skeletonize.

• refine (bool) – If True, refine skeleton nodes by moving them to the center of their corresponding chunk meshes using the L2 cache.

• drop_missing (bool) – Only relevant if refine=True: If True, drop chunks that don’t exist in the L2 cache.

• l2_node_ids (bool) – If True, use the L2 IDs as node IDs.

• omit_failures (bool, optional) – Behaviour when skeleton generation fails. None (default) raises, True skips, False returns empty TreeNeuron.

• progress (bool) – Whether to show a progress bar.

• max_threads (int) – Number of parallel requests to make when fetching the L2 skeletons.

• dataset (str, optional) – Against which CRANTb dataset to query. If None, will use the default dataset.

• **kwargs – Additional keyword arguments passed to TreeNeuron initialization.

Returns:

The extracted L2 skeleton(s).

Return type:

navis.TreeNeuron or navis.NeuronList