PixelReplacement

class jwst.pixel_replace.pixel_replace.PixelReplacement(input_model, **pars)

Bases: object

Main class for performing pixel replacement.

This class controls loading the input data model, selecting the method for pixel replacement, and executing each step. This class should provide modularization to allow for multiple options and possible future reference files.

Parameters:

input_modelJwstDataModel

Datamodel with bad pixels to replace. Updated in-place.

**parsdict, optional

Optional parameters to modify how pixel replacement will execute.

Attributes Summary

DO_NOT_USE
FLUX_ESTIMATED
HORIZONTAL
LOG_SLICE
NON_SCIENCE
VERTICAL

Methods Summary

custom_slice(dispaxis, index)	Construct slice for ease of use with varying dispersion axis.
fit_profile(arrays)	Replace pixels with the profile fit method.
mingrad(arrays)	Replace pixels with the minimum gradient replacement method.
profile_mse(scale, median, current)	Calculate mean-squared error of fitted profile.
replace()	Unpack model and apply pixel replacement algorithm.

Attributes Documentation

DO_NOT_USE = 1

FLUX_ESTIMATED = 268435456

HORIZONTAL = 1

LOG_SLICE = ['column', 'row']

NON_SCIENCE = 512

VERTICAL = 2

Methods Documentation

custom_slice(dispaxis, index)

Construct slice for ease of use with varying dispersion axis.

Parameters:

dispaxisint

Using module-defined:

• 1 = HORIZONTAL

• 2 = VERTICAL

indexint or list

Index or indices of cross-dispersion vectors to slice

Returns:

tuple

Slice constructed using numpy

fit_profile(arrays)

Replace pixels with the profile fit method.

Fit a profile to adjacent columns, scale profile to column with missing pixel(s), and find flux estimate from scaled profile.

Error and variance values for the replaced pixels are similarly estimated, using the scales from the profile fit to the data.

Parameters:

arraysPixelReplaceArrays

Pixel arrays and dispersion direction for the 2D spectrum to process. Arrays are modified in place.

Returns:

arraysPixelReplaceArrays

The input with bad pixels now flagged with FLUX_ESTIMATED and holding a flux value estimated from the spatial profile.

mingrad(arrays)

Replace pixels with the minimum gradient replacement method.

Test the gradient along the spatial and spectral axes using immediately adjacent pixels. Pick whichever dimension has the minimum absolute gradient and replace the missing pixel with the average of the two adjacent pixels along that dimension.

This aims to make the process extremely local; near point sources it should do the replacement along the spectral axis avoiding sampling issues, while near bright extended emission line the replacement should be along the spatial axis. May still be suboptimal near bright emission lines from unresolved point sources.

Does not attempt any replacement if a NaN value is bordered by another NaN value along a given axis.

Parameters:

arraysPixelReplaceArrays

Pixel arrays and dispersion direction for the 2D spectrum to process. Arrays are modified in-place.

Returns:

arraysPixelReplaceArrays

The input with flagged bad pixels now flagged with FLUX_ESTIMATED and holding a flux value estimated from adjacent pixels.

profile_mse(scale, median, current)

Calculate mean-squared error of fitted profile.

Parameters:

scalefloat

Initial estimate of scale factor to bring normalized median profile up to current profile

medianndarray

Median profile constructed from neighboring profile slices

currentndarray

Current profile with bad pixels to be replaced

Returns:

float

Mean-squared error for minimization purposes

replace()

Unpack model and apply pixel replacement algorithm.

Process the input JwstDataModel, unpack any model that holds more than one 2D spectrum, then apply selected algorithm to each 2D spectrum in input.