Source code for pyhiperta.leakage

import numpy as np


[docs] def leakage(waveforms: np.ndarray, intensity: np.ndarray, leakage_mask: np.ndarray) -> np.ndarray: """Compute the leakage 1 and 2, and their intensity. The leakage 1, respectively 2, is the number of pixels that are not 0 after cleaning on the outermost, respectively 2nd outermost, ring of the camera. The leakage intensity is the charge of the leakage pixels, so leakage 1, resp. 2, intensity is the sum of the signal in all pixels of the outermost, resp. 2nd outermost, ring of the camera. Parameters ---------- waveforms : np.ndarray 1D shower image(s). If a single image is provided the shape must be (N_pixels,). If a batch of images is provided, the shape should be (N_batch, N_pixels). Shape: ([N_batch,] N_pixels) intensity : np.ndarray Total charge of each waveform. Shape: ([B_batch,] 1) leakage_mask : np.ndarray Boolean array with value True for pixels on the outermost rings of the camera, False otherwise. leakage_mask[0, :] selects the outermost ring (leakage 1) while leakage_mask[1, :] selects the second outermost ring (leakage 2). Shape: (2, N_pixels) Returns ------- np.ndarray Leakage array. This is a 1D array with shape (4,) if a single image was provided, or a 2D array with shape (N_batch, 4) if a batch of images was provided. The parameters are ordered like so: leakage[..., 0]: leakage 1: number of pixels with signal on the outermost ring of the camera, normalized by the number of pixels with signal. leakage[..., 1]: leakage 2: number of pixels with signal on the 1st and 2nd outermost ring of the camera, normalized by the number of pixels with signal. leakage[..., 3]: leakage 1 intensity: Waveform charge contained in the outermost ring of the camera, normalized by the total charge in the camera leakage[..., 4]: leakage 2 intensity: Waveform charge contained in the 1st and 2nd outermost ring of the camera, normalized by the total charge in the camera """ # number of pixels for normalization nb_pixels = np.count_nonzero(waveforms, axis=-1, keepdims=True).astype(np.float32) # apply leakage mask (both leakage 1 and leakage 2 with broadcasting) leakage_charge = waveforms[..., np.newaxis, :] * leakage_mask # normalized number of pixels with signal in leakage for both leakages leakage = np.count_nonzero(leakage_charge, axis=-1).astype(np.float32) / nb_pixels # normalized leakage intensity for both leakage (add axis to intensity for broadcasting!) leakage_intensity = leakage_charge.sum(axis=-1, dtype=np.float32) / intensity[..., np.newaxis] return np.concatenate([leakage, leakage_intensity], axis=-1)
# return np.stack([leakage[..., 0], leakage[..., 1], leakage_intensity[..., 0], leakage_intensity[..., 1]], axis=-1)