RandomNoisePixelDefect

class cabaret.camera.RandomNoisePixelDefect(name: str = 'defect', rate: float = 0.0, seed: int = 0, apply_before_base: bool = False, _rng: Generator | None = None, _pixels: ndarray | None = None, noise_level: float = 10.0, distribution: Literal['normal', 'poisson'] = 'normal')

A pixel defect that introduces random noise to selected pixels.

This is different from read noise, which is applied to all pixels and represents the electronic noise of the sensor. RandomNoisePixelDefect simulates pixels that are abnormally noisy (e.g., “noisy” or “unstable” pixels), and the noise is added only to those selected pixels.

Examples

>>> from cabaret import Observatory, Sources
>>> pixel_defects = {"noise": {"type": "noise", "rate": 0.1, "noise_level": 5e3}}
>>> observatory = Observatory(camera={"pixel_defects": pixel_defects})
>>> sources = Sources.get_test_sources()
>>> ra, dec = sources.center
>>> _, clean_image, image = observatory.generate_image_stack(
...     exp_time=3, ra=ra, dec=dec, sources=sources, convert_all_to_adu=True
... )

To plot the images, you can use the plot_image function from cabaret.plot:

>>> from cabaret.plot import plot_image
>>> import matplotlib.pyplot as plt
>>> fig, axes = plt.subplots(1, 2, figsize=(7, 5), sharex=True, sharey=True)
>>> _ = plot_image(clean_image, ax=axes[0], title="Image without defects")
>>> _ = plot_image(image, ax=axes[1], title="Image with random noise defects")
>>> plt.subplots_adjust(wspace=0.1)
>>> plt.show()

Methods

__init__([name, rate, seed, ...])
generate_noise(size)	Generate noise for the defect pixels.
introduce_pixel_defect(image, camera[, seed])	Introduce the defect into the image.
number_of_defect_pixels(camera)	Calculate the number of pixels affected by the defect.
set_pixels(pixels, camera)	Set the pixels for the defect.

Attributes

apply_before_base	If True, this defect is applied before the base image (bias, dark, read noise) is added.
distribution	Distribution type for the noise ('normal' or 'poisson').
name	Name of the defect.
noise_level	Standard deviation or scale for the noise.
pixels	Pixel coordinates affected by the defect.
rate	Fraction of pixels affected by the defect.
rng	Random number generator instance.
seed	Random seed for reproducibility.