clustering — Clustering facilities

Module documentation

Clustering facilities.

class hexsample.clustering.Cluster(x: ndarray, y: ndarray, pha: ndarray)

Small container class describing a cluster.

x: ndarray

y: ndarray

pha: ndarray

size() → int

Return the size of the cluster.

pulse_height() → float

Return the total pulse height of the cluster.

centroid() → Tuple[float, float]

Return the cluster centroid.

eta(gamma: float, pitch: float) → Tuple[float, float]

Return the cluster reconstructed position using the eta function.

Arguments

gammafloat

The index of the power law of the eta function.

pitchfloat

The pitch of the pixels.

class hexsample.clustering.ClusteringBase(grid: HexagonalGrid, zero_sup_threshold: float)

Base class for the clustering.

grid: HexagonalGrid

zero_sup_threshold: float

zero_suppress(array: ndarray) → ndarray

Zero suppress a generic array.

run(event: DigiEventRectangular) → Cluster

Workhorse method to be reimplemented by derived classes.

class hexsample.clustering.ClusteringNN(grid: HexagonalGrid, zero_sup_threshold: float, num_neighbors: int)

Neirest neighbor clustering.

This is a very simple clustering strategy where we use the highest pixel in the event as a seed, loop over the six neighbors (after the zero suppression) and keep the N highest pixels.

Arguments

num_neighborsint

The number of neighbors (between 0 and 6) to include in the cluster.

num_neighbors: int

run(event) → Cluster

Overladed method.

Warning

The loop ever the neighbors might likely be vectorized and streamlined for speed using proper numpy array for the offset indexes.