HorneExtract

class specreduce.extract.HorneExtract(image: NDData, trace_object: Trace, bkgrd_prof: None | Model = None, spatial_profile: str | dict = 'gaussian', variance: ndarray = None, mask: ndarray = None, unit: ndarray = None, disp_axis: int = 1, crossdisp_axis: int = 0)

Bases: SpecreduceOperation

Perform a Horne (a.k.a. optimal) extraction on a two-dimensional spectrum.

There are two options for fitting the spatial profile used for extraction - by default, a 1D gaussian is fit and as a uniform profile across the spectrum. Alternativley, the self profile option may be chosen - when this option is chosen, the spatial profile will be sampled (using a default of 10 sample bins, but can be modified with spatial_profile) and interpolated between to produce a smoothly varying spatial profile across the spectrum.

If using the Gaussian option for the spatial profile, a background profile may be fit (but not subtracted) simultaneously to the data. By default, this is done with a 2nd degree polynomial. If using the interpolated_profile option, the background model must be set to None.

Parameters:

imageNDData-like or array-like, required

The input 2D spectrum from which to extract a source. An NDData object must specify uncertainty and a mask. An array requires use of the variance, mask, & unit arguments.

trace_objectTrace, required

The associated 1D trace object created for the 2D image.

disp_axisint, optional

The index of the image’s dispersion axis. [default: 1]

crossdisp_axisint, optional

The index of the image’s cross-dispersion axis. [default: 0]

bkgrd_profModel or None, optional

A model for the image’s background flux when using the gaussian spatial profile. If spatial_profile is set to gaussian, it defaults to models.Polynomial1D(2). Note that the interpolated_profile option does not support a background model, so bkgrd_prof must be left as None.

spatial_profilestr or dict, optional

The shape of the object profile. The first option is ‘gaussian’ to fit a uniform 1D gaussian to the average of pixels in the cross-dispersion direction. The other option is ‘interpolated_profile’ - when this option is used, the profile is sampled in bins and these samples are interpolated between to construct a continuously varying, empirical spatial profile for extraction. For this option, if passed in as a string (i.e spatial_profile=’interpolated_profile’) the default values for the number of bins used (10) and degree of interpolation (linear in x and y, by default) will be used. To set these parameters, pass in a dictionary with the keys ‘n_bins_interpolated_profile’ (which accepts an integer number of bins) and ‘interp_degree’ (which accepts an int, or tuple of ints for x and y degree, respectively). [default: gaussian]

variancendarray, optional

(Only used if image is not an NDData object.) The associated variances for each pixel in the image. Must have the same dimensions as image. If all zeros, the variance will be ignored and treated as all ones. If any zeros, those elements will be excluded via masking. If any negative values, an error will be raised. [default: None]

maskndarray, optional

(Only used if image is not an NDData object.) Whether to mask each pixel in the image. Must have the same dimensions as image. If blank, all non-NaN pixels are unmasked. [default: None]

unitUnit or str, optional

(Only used if image is not an NDData object.) The associated unit for the data in image. If blank, fluxes are interpreted in DN. [default: None]

Attributes Summary

bkgrd_prof
crossdisp_axis
disp_axis
image
mask
spatial_profile
spectrum
trace_object
unit
variance

Methods Summary

__call__([image, trace_object, disp_axis, ...])

Run the Horne calculation on a region of an image and extract a 1D spectrum.

Attributes Documentation

bkgrd_prof: None | Model = None

crossdisp_axis: int = 0

disp_axis: int = 1

image: NDData = <dataclasses._MISSING_TYPE object>

mask: ndarray = None

spatial_profile: str | dict = 'gaussian'

spectrum

trace_object: Trace = <dataclasses._MISSING_TYPE object>

unit: ndarray = None

variance: ndarray = None

Methods Documentation

__call__(image=None, trace_object=None, disp_axis=None, crossdisp_axis=None, bkgrd_prof=None, spatial_profile=None, n_bins_interpolated_profile=None, interp_degree_interpolated_profile=None, variance=None, mask=None, unit=None)

Run the Horne calculation on a region of an image and extract a 1D spectrum.

Parameters:

imageNDData-like or array-like, required

The input 2D spectrum from which to extract a source. An NDData object must specify uncertainty and a mask. An array requires use of the variance, mask, & unit arguments.

trace_objectTrace, required

The associated 1D trace object created for the 2D image.

disp_axisint, optional

The index of the image’s dispersion axis.

crossdisp_axisint, optional

The index of the image’s cross-dispersion axis.

bkgrd_profModel, optional

A model for the image’s background flux when using the gaussian spatial profile. If spatial_profile is set to gaussian, it defaults to models.Polynomial1D(2). Note that the interpolated_profile option does not support a background model, so bkgrd_prof must be left as None.

spatial_profilestr or dict, optional

The shape of the object profile. The first option is ‘gaussian’ to fit a uniform 1D gaussian to the average of pixels in the cross-dispersion direction. The other option is ‘interpolated_profile’ - when this option is used, the profile is sampled in bins and these samples are interpolated between to construct a continuously varying, empirical spatial profile for extraction. For this option, if passed in as a string (i.e spatial_profile=’interpolated_profile’) the default values for the number of bins used (10) and degree of interpolation (linear in x and y, by default) will be used. To set these parameters, pass in a dictionary with the keys ‘n_bins_interpolated_profile’ (which accepts an integer number of bins) and ‘interp_degree’ (which accepts an int, or tuple of ints for x and y degree, respectively). [default: gaussian]

variancendarray, optional

(Only used if image is not an NDData object.) The associated variances for each pixel in the image. Must have the same dimensions as image. If all zeros, the variance will be ignored and treated as all ones. If any zeros, those elements will be excluded via masking. If any negative values, an error will be raised.

maskndarray, optional

(Only used if image is not an NDData object.) Whether to mask each pixel in the image. Must have the same dimensions as image. If blank, all non-NaN pixels are unmasked.

unitUnit or str, optional

(Only used if image is not an NDData object.) The associated unit for the data in image. If blank, fluxes are interpreted in DN.

Returns:

spec_1dSpectrum1D

The final, Horne extracted 1D spectrum.