Integration

class jwst.extract_1d.soss_extract.soss_extract.Integration(scidata, scierr, scimask, refmask, order_models, box_weights, do_bkgsub=True, extract_order3=True, save_intermediate=False, order2_separation_cutoff=(0.77, 0.95))

Bases: object

Class to handle extraction of a single integration.

Parameters:

scidatandarray

2D science data array.

scierrndarray

2D science error array.

scimaskndarray

2D boolean mask array for the science data. True values are masked.

refmaskndarray

2D boolean mask array for the reference pixels. True values are masked.

order_modelslist of DetectorModelOrder

Models of the detector and trace properties, one per spectral order.

box_weightsdict

A dictionary of the weights for each order.

do_bkgsubbool, optional

Whether to perform background subtraction. Default is True.

extract_order3bool, optional

Whether to extract order 3. Default is True.

save_intermediatebool, optional

Whether to save intermediate products for debugging. Default is False.

order2_separation_cutofftuple(float, float), optional

Wavelengths short of which to consider Order 2 to be well-separated from order 1. Two values are defined. The first is the cutoff longwave of which should be included in the combined extraction; the second is the cutoff shortwave of which should be included in the extraction of just that order. Some overlap is helpful to avoid numerical issues.

Methods Summary

decontaminate_image(tracemodels)	Perform decontamination of the image based on the trace models.
estim_flux_first_order([threshold])	Roughly estimate the underlying flux of the target spectrum.
extract_image(decontaminated_data[, ...])	Perform the box-extraction on the image using the trace model to correct for contamination.
make_decontamination_grid(estimate[, rtol, ...])	Create the grid to use for the simultaneous extraction of order 1 and 2.
model_image(wave_grid[, estimate, ...])	Extract the uncontaminated spectra of using the ATOCA algorithm.

Methods Documentation

decontaminate_image(tracemodels)

Perform decontamination of the image based on the trace models.

Parameters:

tracemodelsdict

Dictionary of the modeled detector images for each order.

Returns:

decontaminated_datadict

Dictionary of the decontaminated data for each order.

estim_flux_first_order(threshold=0.0001)

Roughly estimate the underlying flux of the target spectrum.

This is done by simply masking out order 2 and retrieving the flux from order 1.

Parameters:

thresholdfloat, optional

The pixels with an aperture[order 2] > threshold are considered contaminated and will be masked.

Returns:

func

A spline estimator that provides the underlying flux as a function of wavelength

extract_image(decontaminated_data, bad_pix='model', tracemodels=None, verbose=False)

Perform the box-extraction on the image using the trace model to correct for contamination.

Parameters:

decontaminated_datandarray

A single background subtracted NIRISS SOSS detector image.

bad_pixstr

How to handle the bad pixels. Options are ‘masking’ and ‘model’. ‘masking’ will simply mask the bad pixels, such that the number of pixels in each column in the box extraction will not be constant, while the ‘model’ option uses tracemodels to replace the bad pixels.

tracemodelsdict

Dictionary of the modeled detector images for each order.

verbosebool

Print bad pixel imputation messages to INFO-level logging?

Returns:

fluxes, fluxerrs, npixelsdict

Each output is a dictionary, with each extracted order as a key.

make_decontamination_grid(estimate, rtol=0.0001, max_grid_size=20000, n_os=2)

Create the grid to use for the simultaneous extraction of order 1 and 2.

The grid is made by:

1. requiring that it satisfies the oversampling n_os,

2. trying to reach the specified tolerance for the spectral range shared between order 1 and 2, and

3. trying to reach the specified tolerance in the rest of spectral range.

The max_grid_size overrules steps 2 and 3, so the precision may not be reached if the grid size needed is too large.

Parameters:

estimateUnivariateSpline

Estimate of the target flux as a function of wavelength in microns.

rtolfloat, optional

The relative tolerance needed on a pixel model.

max_grid_sizeint, optional

Maximum grid size allowed.

n_osint, optional

The oversampling factor of the wavelength grid used when solving for the uncontaminated flux.

Returns:

wave_gridndarray

The 1D grid of the pixels boundaries at the native sampling.

model_image(wave_grid, estimate=None, tikfacs_in=None, threshold=0.01)

Extract the uncontaminated spectra of using the ATOCA algorithm.

The model consists of three separate extractions:

1. A simultaneous extraction of orders 1 and 2.

2. An extraction of the blue (uncontaminated) part of order 2.

3. An extraction of order 3, if requested.

For each of these extractions, the Tikhonov factor can either be provided using the tikfacs_in dictionary, or else determined via a grid search.

Parameters:

wave_gridndarray or None

The wavelength grid to use for the simultaneous extraction of orders 1 and 2.

estimatefunc, optional

An estimate of the target flux as a function of wavelength in microns. This is used to generate the wavelength grid if wave_grid is None, and to estimate the Tikhonov factor if not provided in tikfacs_in. If wave_grid is given and the Tikhonov factor for order 1 is provided in tikfacs_in, this is not needed. If either is not provided, this must be given.

tikfacs_indict, optional

A dictionary with keys “Order 1”, “Order 2”, and “Order 3” and values giving the Tikhonov factor to use for each order. If None (default), the Tikhonov factor that minimizes the chi-squared for each order will be determined via a grid search.

thresholdfloat, optional

The threshold (between 0 and 1) for determining which pixels to include in the extraction. Pixels with a normalized flux contribution from the spectral trace above this threshold will be included. Default is 1e-2.

Returns:

tracemodelsdict

A dictionary with keys “Order 1”, “Order 2”, and “Order 3” (if extracted) and values giving the modeled detector image for each order.

spec_listlist of SpecModel

A list of the extracted spectra for each order. If the Tikhonov factor was determined via a grid search, this will include the intermediate spectra for each factor tested, with the best-fitting spectrum last in the list.

tikfacs_outdict

A dictionary with keys “Order 1”, “Order 2”, and “Order 3” (if extracted) and values giving the Tikhonov factor used for each order.