SynthImage#

class specreduce.utils.synth_data.SynthImage(nx: int = 3000, ny: int = 1000, wcs: WCS | None = None, extent=(3500, 7000), wave_unit: Unit = Unit('Angstrom'), seed: int | None = None)[source]#

Bases: object

Immutable, composable builder for synthetic 2D spectroscopic images.

Build an image by chaining add_* methods, then render it with one of the to_* terminal methods. Each add_* returns a new SynthImage; the original is never mutated, so a base configuration can be safely branched.

Parameters:

nx: Size of the image along the X (dispersion) axis.
ny: Size of the image along the Y (spatial) axis.
wcs: Optional 2D WCS with a single spectral axis. If not provided and arc layers are present, a linear WAVE/PIXEL WCS is built from extent and wave_unit.
extent: Beginning and end wavelengths used to build a default WCS when wcs is not supplied and arc layers are present.
wave_unit: Wavelength unit for the default WCS.
seed: Seed for the random number generator used by the noise layers. If None, noise is non-deterministic.

Examples

Build a traced continuum source with background and noise, then render it:

from astropy.modeling import models
from specreduce.utils.synth_data import SynthImage

image = (
    SynthImage(nx=1024, ny=400, seed=42)
    .add_background(5)
    .add_source(profile=models.Moffat1D(amplitude=20, alpha=0.1))
    .add_poisson_noise()
    .add_read_noise(3)
    .to_ccddata()
)

See the Synthetic Data guide for more examples, including tilted arc lines and modeling a non-linear dispersion relation.

Methods Summary

`add_arcs`([linelists, line_fwhm, ...])	Add emission lines from one or more pypeit calibration line lists.
`add_background`(level)	Add a constant background level (counts).
`add_poisson_noise`()	Apply Poisson noise to the rendered signal (requires photutils).
`add_rdnoise`(sigma)	Add Gaussian read noise of standard deviation `sigma` (counts).
`add_read_noise`(sigma)	Add Gaussian read noise of standard deviation `sigma` (counts).
`add_skylines`([linelists])	Add night-sky airglow emission lines (OH lists), wrapping `add_arcs`.
`add_source`([profile, trace_center, ...])	Add a source with a Chebyshev-traced spatial profile.
`to_array`()	Render and return the image as a plain `numpy.ndarray` (counts).
`to_ccddata`()	Render and return the image as a `CCDData`.
`to_spectrum`()	Render and return the image as a `Spectrum`.

Methods Documentation

add_arcs(linelists=('HeI',), line_fwhm: float = 5.0, amplitude_scale: float = 1.0, wave_air: bool = False, tilt_func: Model = None) → SynthImage[source]#: Add emission lines from one or more pypeit calibration line lists.

add_background(level: float) → SynthImage[source]#: Add a constant background level (counts).

add_poisson_noise() → SynthImage[source]#: Apply Poisson noise to the rendered signal (requires photutils).

add_rdnoise(sigma: float) → SynthImage#: Add Gaussian read noise of standard deviation sigma (counts).

add_read_noise(sigma: float) → SynthImage[source]#: Add Gaussian read noise of standard deviation sigma (counts).

add_skylines(linelists='OH_GMOS', **kwargs) → SynthImage[source]#: Add night-sky airglow emission lines (OH lists), wrapping add_arcs.

add_source(profile: Model = None, trace_center: float | None = None, trace_order: int = 3, trace_coeffs: dict | None = None, spectrum: Spectrum = None) → SynthImage[source]#

Add a source with a Chebyshev-traced spatial profile.

Parameters:

profile: Astropy model describing the cross-dispersion spatial profile of the source. Defaults to Moffat1D(amplitude=10, alpha=0.1).
trace_center: Central cross-dispersion position of the trace. Defaults to the middle of the image.
trace_order: Polynomial order of the Chebyshev trace.
trace_coeffs: Coefficients of the Chebyshev trace, e.g. {"c0": 0, "c1": 50}.
spectrum: Optional 1D Spectrum describing the dispersion-axis flux of the source. Its flux is resampled onto the image wavelength grid (requiring a resolvable WCS) and normalized so its peak within the image extent is one. Wavelengths outside the spectrum’s range are set to zero. The normalized flux multiplies the spatial profile column by column, so the source amplitude varies with wavelength. When None (default) the source has a flat continuum.

to_array() → ndarray[source]#: Render and return the image as a plain numpy.ndarray (counts).

to_ccddata() → CCDData[source]#: Render and return the image as a CCDData.

to_spectrum() → Spectrum[source]#: Render and return the image as a Spectrum.