Storable Data Types

class larcv.Particle

A Particle represents an abstract label more relevant to experimental data. Typically created from simulation data, there are fields to hold a variety of per-particle information as well as reconstructing hierarchy information.

Technically, for storage reasons, there is an underlying “ParticleHolder” that stores the data in a contiguous memory region with appropriate data types. You can access this object directly by calling as_array and get a numpy structured array data object.

ancestor_pdg_code(*args, **kwargs)

Overloaded function.

  1. ancestor_pdg_code(self: larcv.pylarcv.Particle) -> int

Get the value of ancestor_pdg_code for this particle.

  1. ancestor_pdg_code(self: larcv.pylarcv.Particle, ancestor_pdg_code: int) -> None

Set the value of ancestor_pdg_code for this particle.

ancestor_position(*args, **kwargs)

Overloaded function.

  1. ancestor_position(self: larcv.pylarcv.Particle) -> larcv.pylarcv.Vertex

    Return the ancestor position of this particle.

  2. ancestor_position(self: larcv.pylarcv.Particle, vertex: larcv.pylarcv.Vertex) -> None

    Set the ancestor position of this particle.

  3. ancestor_position(self: larcv.pylarcv.Particle, x: float, y: float, z: float, t: float) -> None

    Set the ancestor position of this particle.

ancestor_t(self: larcv.pylarcv.Particle) → float

Get the ancestor_t of this particle.

ancestor_track_id(*args, **kwargs)

Overloaded function.

  1. ancestor_track_id(self: larcv.pylarcv.Particle) -> int

Get the value of ancestor_track_id for this particle.

  1. ancestor_track_id(self: larcv.pylarcv.Particle, ancestor_track_id: int) -> None

Set the value of ancestor_track_id for this particle.

ancestor_x(self: larcv.pylarcv.Particle) → float

Get the ancestor_x of this particle.

ancestor_y(self: larcv.pylarcv.Particle) → float

Get the ancestor_y of this particle.

ancestor_z(self: larcv.pylarcv.Particle) → float

Get the ancestor_z of this particle.

as_array(self: larcv.pylarcv.Particle) → numpy.ndarray[larcv3::ParticleHolder]

Return the underlying ParticleHolder object of this particle as a numpy structured array.

creation_process(*args, **kwargs)

Overloaded function.

  1. creation_process(self: larcv.pylarcv.Particle) -> str

Get the value of creation_process for this particle.

  1. creation_process(self: larcv.pylarcv.Particle, creation_process: str) -> None

Set the value of creation_process for this particle.

distance_travel(*args, **kwargs)

Overloaded function.

  1. distance_travel(self: larcv.pylarcv.Particle) -> float

Get the value of distance_travel for this particle.

  1. distance_travel(self: larcv.pylarcv.Particle, distance_travel: float) -> None

Set the value of distance_travel for this particle.

dump(self: larcv.pylarcv.Particle) → str

Return a string representation for this particle.

end_position(*args, **kwargs)

Overloaded function.

  1. end_position(self: larcv.pylarcv.Particle) -> larcv.pylarcv.Vertex

Get the value of end_position for this particle.

  1. end_position(self: larcv.pylarcv.Particle, end_position: larcv.pylarcv.Vertex) -> None

Set the value of end_position with a Vertex object.

  1. end_position(self: larcv.pylarcv.Particle, x: float, y: float, z: float, t: float) -> None

Set the value of end_position with x, y, z, and t.

energy_deposit(*args, **kwargs)

Overloaded function.

  1. energy_deposit(self: larcv.pylarcv.Particle) -> float

Get the value of energy_deposit for this particle.

  1. energy_deposit(self: larcv.pylarcv.Particle, energy_deposit: float) -> None

Set the value of energy_deposit for this particle.

energy_init(*args, **kwargs)

Overloaded function.

  1. energy_init(self: larcv.pylarcv.Particle) -> float

Get the value of energy_init for this particle.

  1. energy_init(self: larcv.pylarcv.Particle, energy_init: float) -> None

Set the value of energy_init for this particle.

first_step(*args, **kwargs)

Overloaded function.

  1. first_step(self: larcv.pylarcv.Particle) -> larcv.pylarcv.Vertex

Get the value of first_step for this particle.

  1. first_step(self: larcv.pylarcv.Particle, first_step: larcv.pylarcv.Vertex) -> None

Set the value of first_step with a Vertex object.

  1. first_step(self: larcv.pylarcv.Particle, x: float, y: float, z: float, t: float) -> None

Set the value of first_step with x, y, z, and t.

id(*args, **kwargs)

Overloaded function.

  1. id(self: larcv.pylarcv.Particle) -> int

Get the id for this particle.

  1. id(self: larcv.pylarcv.Particle, id: int) -> None

Set the id for this particle.

last_step(*args, **kwargs)

Overloaded function.

  1. last_step(self: larcv.pylarcv.Particle) -> larcv.pylarcv.Vertex

Get the value of last_step for this particle.

  1. last_step(self: larcv.pylarcv.Particle, last_step: larcv.pylarcv.Vertex) -> None

Set the value of last_step with a Vertex object.

  1. last_step(self: larcv.pylarcv.Particle, x: float, y: float, z: float, t: float) -> None

Set the value of last_step with x, y, z, and t.

mcst_index(*args, **kwargs)

Overloaded function.

  1. mcst_index(self: larcv.pylarcv.Particle) -> int

Get the mcst_index for this particle.

  1. mcst_index(self: larcv.pylarcv.Particle, mcst_index: int) -> None

Set the mcst_index for this particle.

mct_index(*args, **kwargs)

Overloaded function.

  1. mct_index(self: larcv.pylarcv.Particle) -> int

Get the mct_index for this particle.

  1. mct_index(self: larcv.pylarcv.Particle, mct_index: int) -> None

Set the mct_index for this particle.

momentum(self: larcv.pylarcv.Particle, arg0: float, arg1: float, arg2: float) → None

Get the momentum of this particle.

nu_current_type(*args, **kwargs)

Overloaded function.

  1. nu_current_type(self: larcv.pylarcv.Particle) -> int

Get the nu_current_type for this particle.

  1. nu_current_type(self: larcv.pylarcv.Particle, nu_current_type: int) -> None

Set the nu_current_type for this particle.

nu_interaction_type(*args, **kwargs)

Overloaded function.

  1. nu_interaction_type(self: larcv.pylarcv.Particle) -> int

Get the nu_interaction_type for this particle.

  1. nu_interaction_type(self: larcv.pylarcv.Particle, nu_interaction_type: int) -> None

Set the nu_interaction_type for this particle.

p(self: larcv.pylarcv.Particle) → float

Get the p of this particle.

parent_pdg_code(*args, **kwargs)

Overloaded function.

  1. parent_pdg_code(self: larcv.pylarcv.Particle) -> int

Get the value of parent_pdg_code for this particle.

  1. parent_pdg_code(self: larcv.pylarcv.Particle, parent_pdg_code: int) -> None

Set the value of parent_pdg_code for this particle.

parent_position(*args, **kwargs)

Overloaded function.

  1. parent_position(self: larcv.pylarcv.Particle) -> larcv.pylarcv.Vertex

Get the value of parent_position for this particle.

  1. parent_position(self: larcv.pylarcv.Particle, parent_position: larcv.pylarcv.Vertex) -> None

Set the value of parent_position for this particle.

  1. parent_position(self: larcv.pylarcv.Particle, x: float, y: float, z: float, t: float) -> None

Set the value of parent_position with x, y, z, and t.

parent_t(self: larcv.pylarcv.Particle) → float

Get the parent_t of this particle.

parent_track_id(*args, **kwargs)

Overloaded function.

  1. parent_track_id(self: larcv.pylarcv.Particle) -> int

Get the value of parent_track_id for this particle.

  1. parent_track_id(self: larcv.pylarcv.Particle, parent_track_id: int) -> None

Set the value of parent_track_id for this particle.

parent_x(self: larcv.pylarcv.Particle) → float

Get the parent_x of this particle.

parent_y(self: larcv.pylarcv.Particle) → float

Get the parent_y of this particle.

parent_z(self: larcv.pylarcv.Particle) → float

Get the parent_z of this particle.

pdg_code(*args, **kwargs)

Overloaded function.

  1. pdg_code(self: larcv.pylarcv.Particle) -> int

Get the pdg_code for this particle.

  1. pdg_code(self: larcv.pylarcv.Particle, pdg_code: int) -> None

Set the pdg_code for this particle.

position(*args, **kwargs)

Overloaded function.

  1. position(self: larcv.pylarcv.Particle) -> larcv.pylarcv.Vertex

  2. position(self: larcv.pylarcv.Particle, vertex: larcv.pylarcv.Vertex) -> None

Set the position of the particle with a Vertex object.

  1. position(self: larcv.pylarcv.Particle, x: float, y: float, z: float, t: float) -> None

Set the position of the particle with x, y, z, t object.

px(self: larcv.pylarcv.Particle) → float

Get the px of this particle.

py(self: larcv.pylarcv.Particle) → float

Get the py of this particle.

pz(self: larcv.pylarcv.Particle) → float

Get the pz of this particle.

shape(*args, **kwargs)

Overloaded function.

  1. shape(self: larcv.pylarcv.Particle) -> larcv3::ShapeType_t

Get the shape for this particle.

  1. shape(self: larcv.pylarcv.Particle, shape: larcv3::ShapeType_t) -> None

Set the shape for this particle.

t(self: larcv.pylarcv.Particle) → float

Get the t of this particle.

track_id(*args, **kwargs)

Overloaded function.

  1. track_id(self: larcv.pylarcv.Particle) -> int

Get the track_id for this particle.

  1. track_id(self: larcv.pylarcv.Particle, track_id: int) -> None

Set the track_id for this particle.

x(self: larcv.pylarcv.Particle) → float

Get the x of this particle.

y(self: larcv.pylarcv.Particle) → float

Get the y of this particle.

z(self: larcv.pylarcv.Particle) → float

Get the z of this particle.

Bounding Boxes are available in 2D and 3D. Only the 2D api is shown here for brevity, but the function calls are identical in 3D.

class larcv.BBox2D

Represents a bounding box in N dimensions, defined by a centroid and half_length in each axis from the centroid. The rotation component is currently not used but provided in the data store implementation for future use. Used to represent, for example, the location of a particle interaction, a bounding box for an instance of an object, etc.

centroid(self: larcv.pylarcv.BBox2D) → List[float[2]]

Get the value of the centroid.

dump(self: larcv.pylarcv.BBox2D) → str

Return a string representation of the BBox

half_length(self: larcv.pylarcv.BBox2D) → List[float[2]]

Get the value of the half_length.

identity_rotation(self: larcv.pylarcv.BBox2D) → List[float[4]]

Get the value of the identity rotation.

rotation_matrix(self: larcv.pylarcv.BBox2D) → List[float[4]]

Get the value of the rotation matrix.

(Dense) Tensors are available in 1D, 2D, 3D and 4D. Only the 2D api is shown here for brevity, but the function calls are identical in other dimensions.

class larcv.Tensor2D

A Tensor is a densely stored, memory-contiguous representation of image-like data. Tensors have ImageMeta associated to them to provide external context for the image. In this case, the coordinates of a pixel (X/Y/Z …) map to a 1D index through the ImageMeta interface, and are identical to the indexing required in various array interfaces.

For example, a set of coordinates in 2D (for example) x and y might map to an index through the ImageMeta object. In that case, the following techniques should provide identical data:

>>> # Access directly:
>>> tensor.pixel((x,y)) # use the pixel method of the tensor.

>>> # Same as numpy:
>>> tensor.as_array[x][y] # convert this tensor to numpy and access with numpy's array indexing.

>>> # Using indexes:
>>> index = tensor.meta().index((x,y)) # get the index from the image meta
>>> tensor.pixel(index) # access the pixel via the index

>>> # Using numpy methods:
>>> dims = tensor.meta.number_of_voxels()
>>> index = numpy.ravel_multi_index((x,y), dims)
>>> tensor.pixel(index) # use numpy constructed index to get the data directly

The availablitity of both interfaces is to serve as a bridge between numpy.ndarray and SparseTensors, which have only the index-style access (through Voxel IDs).

as_array(self: larcv.pylarcv.Tensor2D) → numpy.ndarray[numpy.float32]

Return the numpy view of this tensor.

as_vector(self: larcv.pylarcv.Tensor2D) → List[float]

Get the single-dimensioned array of memory-contiguous data.

binarize(self: larcv.pylarcv.Tensor2D, threshold: float, lower: float, upper: float) → None

Replace all pixels with either the lower or upper value, based on threshold.

clear_data(self: larcv.pylarcv.Tensor2D) → None

Set the image data to 0.

compress(*args, **kwargs)

Overloaded function.

  1. compress(self: larcv.pylarcv.Tensor2D, compression: List[int[2]], pooltype: larcv.pylarcv.PoolType_t) -> larcv.pylarcv.Tensor2D

Return a new tensor that is compressed/downsampled according to the PoolType, with compression unique per axis.

  1. compress(self: larcv.pylarcv.Tensor2D, compression: int, pool_type: larcv.pylarcv.PoolType_t) -> larcv.pylarcv.Tensor2D

Return a new tensor that is compressed/downsampled equally along all axes.

eltwise(*args, **kwargs)

Overloaded function.

  1. eltwise(self: larcv.pylarcv.Tensor2D, other: larcv.pylarcv.Tensor2D) -> None

Element wise multiplication with another array. Equivalent to dot product in 1D.

  1. eltwise(self: larcv.pylarcv.Tensor2D, other: List[float], allow_longer: bool = False) -> None

Element wise multiplication with another array. Equivalent to dot product in 1D.

meta(self: larcv.pylarcv.Tensor2D) → larcv.pylarcv.ImageMeta2D

Get the meta of this tensor.

paint(self: larcv.pylarcv.Tensor2D, value: float) → None

Set ALL pixels to the specified value.

pixel(*args, **kwargs)

Overloaded function.

  1. pixel(self: larcv.pylarcv.Tensor2D, coordinates: List[int]) -> float

Access a pixel via a list of coordinates.

  1. pixel(self: larcv.pylarcv.Tensor2D, index: int) -> float

Access the pixel at a specified index.

reset(self: larcv.pylarcv.Tensor2D, arg0: larcv.pylarcv.ImageMeta2D) → None

Reset the Tensor to an empty state.

set_pixel(*args, **kwargs)

Overloaded function.

  1. set_pixel(self: larcv.pylarcv.Tensor2D, coordinates: List[int], value: float) -> None

Set an individual pixel’s value at coordinates.

  1. set_pixel(self: larcv.pylarcv.Tensor2D, index: int, value: float) -> None

Set an individual pixel’s value at index.

set_projection_id(self: larcv.pylarcv.Tensor2D, id: int) → None

Set the projection id of the meta for this tensor to id.

size(self: larcv.pylarcv.Tensor2D) → int

Return the total number of points in the data array.

threshold(self: larcv.pylarcv.Tensor2D, threshold: float, lower: bool = True) → None

If lower is true, this is minimum threshold. Otherwise, this is a maximum threshold.

Sparse Tensors are available in 2D and 3D. Only the 2D api is shown here for brevity, but the function calls are identical in 3D.

class larcv.SparseTensor2D

A SparseTensor is a VoxelSet + an ImageMeta to map voxel ids to spatial locations. Available in 2D and 3D.

SparseTensor inherits from VoxelSet and has an ImageMeta attribute, so any methods available in VoxelSet to manipulate the voxels are available here.

This can be used for storing sparse image data, in 2D or 3D, include segmentation labels.

clear_data(self: larcv.pylarcv.SparseTensor2D) → None

Clear all voxels and reset the ImageMeta.

compress(*args, **kwargs)

Overloaded function.

  1. compress(self: larcv.pylarcv.SparseTensor2D, compression: List[int[2]], pooltype: larcv.pylarcv.PoolType_t) -> larcv.pylarcv.SparseTensor2D

Return a new sparse tensor that is compressed/downsampled according to the PoolType, with compression unique per axis.

  1. compress(self: larcv.pylarcv.SparseTensor2D, compression: int, pooltype: larcv.pylarcv.PoolType_t) -> larcv.pylarcv.SparseTensor2D

Return a new sparse tensor that is compressed/downsampled equally along all axes.

dense(self: larcv.pylarcv.SparseTensor2D) → numpy.ndarray[numpy.float32]

Convert this data to a dense representation and return as a numpy.ndarray.

emplace(self: larcv.pylarcv.SparseTensor2D, arg0: larcv.pylarcv.Voxel, arg1: bool) → None

Add a new voxel.

meta(*args, **kwargs)

Overloaded function.

  1. meta(self: larcv.pylarcv.SparseTensor2D) -> larcv.pylarcv.ImageMeta2D

Get the meta for this SparseTensor.

  1. meta(self: larcv.pylarcv.SparseTensor2D, meta: larcv.pylarcv.ImageMeta2D, check: bool = True) -> None

Set the image meta, optionally check if it is valid (default true.

set(self: larcv.pylarcv.SparseTensor2D, voxel_set: larcv.pylarcv.VoxelSet, meta: larcv.pylarcv.ImageMeta2D) → None

Set the voxel_set and the meta simultaneously.

to_tensor(self: larcv.pylarcv.SparseTensor2D) → larcv.pylarcv.Tensor2D

Return the data in this SparseTensor as a corresponding dense Tensor.

Sparse Clusters are available in 2D and 3D. Only the 2D api is shown here for brevity, but the function calls are identical in 3D.

class larcv.SparseCluster2D

A SparseCluster is a VoxelSetArray + an ImageMeta to map voxel ids to spatial locations. Available in 2D and 3D.

SparseCluster inherits from VoxelSetArray and has an ImageMeta attribute, so any methods available in VoxelSetArray to manipulate the voxels are available here.

In a way, SparseCluster is aset of SparseTensors that all share an ImageMeta. This can be used for instance segmentation labels in particular.

clear_data(self: larcv.pylarcv.SparseCluster2D) → None

Clear all data

meta(*args, **kwargs)

Overloaded function.

  1. meta(self: larcv.pylarcv.SparseCluster2D) -> larcv.pylarcv.ImageMeta2D

Get the meta

  1. meta(self: larcv.pylarcv.SparseCluster2D, arg0: larcv.pylarcv.ImageMeta2D) -> None

Set the meta