spine.data.out.RecoParticle

class spine.data.out.RecoParticle(is_truth: bool = False, units: str = 'cm', id: int = -1, is_contained: bool = False, is_time_contained: bool = False, is_cathode_crosser: bool = False, is_matched: bool = False, cathode_offset: float = nan, index: ndarray = <factory>, orig_index: ndarray = <factory>, points: ndarray = <factory>, depositions: ndarray = <factory>, sources: ndarray = <factory>, match_ids: ndarray = <factory>, match_overlaps: ndarray = <factory>, interaction_id: int = -1, chi2_pid: int = -1, is_primary: bool = False, is_crt_matched: bool = False, is_valid: bool = True, length: float = nan, calo_ke: float = nan, csda_ke: float = nan, mcs_ke: float = nan, shape: int = -1, pid: int = -1, fragment_ids: ndarray = <factory>, start_point: ndarray = <factory>, end_point: ndarray = <factory>, chi2_per_pid: ndarray = <factory>, csda_ke_per_pid: ndarray = <factory>, mcs_ke_per_pid: ndarray = <factory>, crt_ids: ndarray = <factory>, crt_times: ndarray = <factory>, crt_scores: ndarray = <factory>, start_dedx: float = nan, end_dedx: float = nan, vertex_distance: float = nan, start_straightness: float = nan, directional_spread: float = nan, axial_spread: float = nan, fragments: list[RecoFragment] = <factory>, start_dir: ndarray = <factory>, end_dir: ndarray = <factory>, pid_scores: ndarray = <factory>, primary_scores: ndarray = <factory>, ppn_ids: ndarray = <factory>, ppn_points: ndarray = <factory>)[source]

Reconstructed particle information.

fragments

List of fragments that make up this particle

Type:: List[RecoFragment]

pdg_code

PDG code corresponding to the PID number

Type:: int

mass

Rest mass of the particle in MeV/c^2

Type:: float

ke

Best guess kinetic energy of the particle in MeV

Type:: float

momentum

3-momentum of the particle at the production point in MeV/c

Type:: np.ndarray

start_dir

Particle direction w.r.t. the start point

Type:: np.ndarray

end_dir

(3) Particle direction w.r.t. the end point (only assigned to track objects)

Type:: np.ndarray

pid_scores

Array of softmax scores associated with each particle class

Type:: np.ndarray

primary_scores

Array of softmax scores associated with secondary and primary

Type:: np.ndarray

ppn_ids

List of indexes of PPN points associated with this particle

Type:: np.ndarray

ppn_points

(M, 3) List of PPN points tagged to this particle

Type:: np.ndarray

start_dedx

dE/dx around a user-defined neighborhood of the start point in MeV/cm

Type:: float

end_dedx

dE/dx around a user-defined neighborhood of the end point in MeV/cm

Type:: float

vertex_distance

Set-to-point distance between all particle points and the parent interaction vertex position in cm

Type:: float

start_straightness

Explained variance ratio of the beginning of the particle

Type:: float

directional_spread

Estimate of the angular spread of the particle (cosine spread)

Type:: float

axial_spread

Pearson correlation coefficient of the axial profile of the particle w.r.t. to the distance from its start point

Type:: float

Attributes:

depositions_sum: Total deposition value for the entire object.
enum_attrs: Fetches HDF5-friendly enumerator descriptors for enum-backed fields.
enum_dicts: Fetches the dictionary of enumerated attributes and their enumerator descriptors.
enum_values: Fetches reverse enum lookups from stored values to symbolic names.
field_units: Fetches the documented units for each field.
index_attrs: Return the tuple of index-bearing attributes.
ke: Best-guess kinetic energy in MeV.
mass: Rest mass of the particle in MeV/c^2.
module_ids: List of modules that contribute to this object.
momentum: Best-guess momentum in MeV/c.
num_fragments: Number of fragments that make up this particle.
p: Computes the magnitude of the initial momentum.
pdg_code: Translates the enumerated particle type to a sign-less PDG code.
reco_end_dir: Alias for end_dir, to match nomenclature in truth.
reco_ke: Alias for ke, to match nomenclature in truth.
reco_length: Alias for length, to match nomenclature in truth.
reco_momentum: Alias for momentum, to match nomenclature in truth.
reco_start_dir: Alias for start_dir, to match nomenclature in truth.
size: Total number of voxels that make up the object.

Methods

`as_dict`([lite, include_derived])	Returns the data class as dictionary of (key, value) pairs.
`attr_names`([include_derived, ...])	Return the names of valid attributes on this data class.
`from_dict`(cls_dict)	Builds and returns an object of the class from a dictionary of attributes.
`merge`(other)	Merge another particle instance into this one.
`reset_cathode_crosser`()	Resets the cathode crossing information for the object.
`reset_crt_match`()	Reset all the CRT hit matching attributes.
`reset_match`()	Resets the reco/truth matching information for the object.
`scalar_dict`([attrs, lengths, lite])	Returns the data class attributes as a dictionary of scalars.
`set_precision`([precision])	Casts all the vector attributes to a different precision.
`shift_indexes`(shifts)	Apply offsets to index attributes in place.
`to_cm`(meta)	Converts the coordinates of the positional attributes to cm in place.
`to_px`(meta)	Converts the coordinates of the positional attributes to pixel in place.
`value_with_units`(attr)	Fetch an attribute value with its documented units.

__init__(is_truth: bool = False, units: str = 'cm', id: int = -1, is_contained: bool = False, is_time_contained: bool = False, is_cathode_crosser: bool = False, is_matched: bool = False, cathode_offset: float = nan, index: ndarray = <factory>, orig_index: ndarray = <factory>, points: ndarray = <factory>, depositions: ndarray = <factory>, sources: ndarray = <factory>, match_ids: ndarray = <factory>, match_overlaps: ndarray = <factory>, interaction_id: int = -1, chi2_pid: int = -1, is_primary: bool = False, is_crt_matched: bool = False, is_valid: bool = True, length: float = nan, calo_ke: float = nan, csda_ke: float = nan, mcs_ke: float = nan, shape: int = -1, pid: int = -1, fragment_ids: ndarray = <factory>, start_point: ndarray = <factory>, end_point: ndarray = <factory>, chi2_per_pid: ndarray = <factory>, csda_ke_per_pid: ndarray = <factory>, mcs_ke_per_pid: ndarray = <factory>, crt_ids: ndarray = <factory>, crt_times: ndarray = <factory>, crt_scores: ndarray = <factory>, start_dedx: float = nan, end_dedx: float = nan, vertex_distance: float = nan, start_straightness: float = nan, directional_spread: float = nan, axial_spread: float = nan, fragments: list[RecoFragment] = <factory>, start_dir: ndarray = <factory>, end_dir: ndarray = <factory>, pid_scores: ndarray = <factory>, primary_scores: ndarray = <factory>, ppn_ids: ndarray = <factory>, ppn_points: ndarray = <factory>) → None

Methods

`__init__`(is_truth, units, id, is_contained, ...)
`as_dict`([lite, include_derived])	Returns the data class as dictionary of (key, value) pairs.
`attr_names`([include_derived, ...])	Return the names of valid attributes on this data class.
`from_dict`(cls_dict)	Builds and returns an object of the class from a dictionary of attributes.
`merge`(other)	Merge another particle instance into this one.
`reset_cathode_crosser`()	Resets the cathode crossing information for the object.
`reset_crt_match`()	Reset all the CRT hit matching attributes.
`reset_match`()	Resets the reco/truth matching information for the object.
`scalar_dict`([attrs, lengths, lite])	Returns the data class attributes as a dictionary of scalars.
`set_precision`([precision])	Casts all the vector attributes to a different precision.
`shift_indexes`(shifts)	Apply offsets to index attributes in place.
`to_cm`(meta)	Converts the coordinates of the positional attributes to cm in place.
`to_px`(meta)	Converts the coordinates of the positional attributes to pixel in place.
`value_with_units`(attr)	Fetch an attribute value with its documented units.

Attributes

`axial_spread`
`calo_ke`
`cathode_offset`
`chi2_pid`
`csda_ke`
`depositions_sum`	Total deposition value for the entire object.
`directional_spread`
`end_dedx`
`enum_attrs`	Fetches HDF5-friendly enumerator descriptors for enum-backed fields.
`enum_dicts`	Fetches the dictionary of enumerated attributes and their enumerator descriptors.
`enum_values`	Fetches reverse enum lookups from stored values to symbolic names.
`field_units`	Fetches the documented units for each field.
`id`
`index_attrs`	Return the tuple of index-bearing attributes.
`interaction_id`
`is_cathode_crosser`
`is_contained`
`is_crt_matched`
`is_matched`
`is_primary`
`is_time_contained`
`is_truth`
`is_valid`
`ke`	Best-guess kinetic energy in MeV.
`length`
`mass`	Rest mass of the particle in MeV/c^2.
`mcs_ke`
`module_ids`	List of modules that contribute to this object.
`momentum`	Best-guess momentum in MeV/c.
`num_fragments`	Number of fragments that make up this particle.
`p`	Computes the magnitude of the initial momentum.
`pdg_code`	Translates the enumerated particle type to a sign-less PDG code.
`pid`
`reco_end_dir`	Alias for end_dir, to match nomenclature in truth.
`reco_ke`	Alias for ke, to match nomenclature in truth.
`reco_length`	Alias for length, to match nomenclature in truth.
`reco_momentum`	Alias for momentum, to match nomenclature in truth.
`reco_start_dir`	Alias for start_dir, to match nomenclature in truth.
`shape`
`size`	Total number of voxels that make up the object.
`start_dedx`
`start_straightness`
`units`
`vertex_distance`
`fragments`
`start_dir`
`end_dir`
`pid_scores`
`primary_scores`
`ppn_ids`
`ppn_points`
`fragment_ids`
`start_point`
`end_point`
`chi2_per_pid`
`csda_ke_per_pid`
`mcs_ke_per_pid`
`crt_ids`
`crt_times`
`crt_scores`
`index`
`orig_index`
`points`
`depositions`
`sources`
`match_ids`
`match_overlaps`

start_dedx: float = nan

end_dedx: float = nan

vertex_distance: float = nan

start_straightness: float = nan

directional_spread: float = nan

axial_spread: float = nan

fragments: list[RecoFragment]

start_dir: ndarray

end_dir: ndarray

pid_scores: ndarray

primary_scores: ndarray

ppn_ids: ndarray

ppn_points: ndarray

merge(other)[source]

Merge another particle instance into this one.

The merging strategy differs depending on the the particle shapes merged together:

Track plus track: the merged particle uses the pair of end points that are farthest apart, and the primary and PID scores are copied from the constituent with the highest primary score.
Shower plus track: the track is merged into the shower, the shower start point is updated to the track end point farthest from the current shower start point, the primary score follows the most primary-like constituent and the PID stays that of the shower.

Parameters:: other (RecoParticle) – Other reconstructed particle to merge into this one

property pdg_code: int

Translates the enumerated particle type to a sign-less PDG code.

Returns:: Reconstructed sign-less PDG code
Return type:: int

property mass: float

Rest mass of the particle in MeV/c^2.

The mass is inferred from the predicted mass.

Returns:: Rest mass of the particle
Return type:: float

property ke: float

Best-guess kinetic energy in MeV.

Uses calorimetry for EM activity and this order for track: - CSDA-based estimate if it is available - MCS-based estimate if it is available - Calorimetry if all else fails

Returns:: Best-guess kinetic energy
Return type:: float

property momentum: ndarray

Best-guess momentum in MeV/c.

Returns:

Momentum vector

Return type:

np.ndarray

property p: float

Computes the magnitude of the initial momentum.

Returns:: Norm of the initial momentum vector
Return type:: float

property reco_ke: float: Alias for ke, to match nomenclature in truth.

property reco_momentum: ndarray: Alias for momentum, to match nomenclature in truth.

property reco_length: float: Alias for length, to match nomenclature in truth.

property reco_start_dir: ndarray: Alias for start_dir, to match nomenclature in truth.

property reco_end_dir: ndarray: Alias for end_dir, to match nomenclature in truth.