romanimpreprocess.utils.maskhandling

Tools for building a mask by growing the input bitmask by different amounts.

Classes

CombinedMask

Class to generate a boolean mask from multiple flags.

Attributes

PixelMask1

Classes

CombinedMask

Class to generate a boolean mask from multiple flags.

Module Contents

class CombinedMask(maskdict)

Class to generate a boolean mask from multiple flags.

Parameters:

maskdict (dict) – A dictionary of how much to grow each flag (see notes for description).

array

Dictionary of how much to grow each flag.

Type:

dict

__init__()

Constructor.

build()

Make boolean mask from a data quality array.

convert_file()

Stand-alone function to make a mask from a file.

Notes

To take a dq array, and flag any pixel with 'gw_affected_data' and the cardinal-nearest neigbors if 'jump_det' is set:

myMaskFunc = CombinedMask({'jump_det': 5, 'gw_affected_data': 1})
mymask = myMaskFunc.build(dq)

The mask names are as described in roman_datamodels.dqflags.pixel. Note that capitalization is automatic so this function is not case-sensitive.

Options for growing are specified by the number of pixels affected: * 1 = that pixel * 5 = cardinal nearest neighbors * 9 = 3x3 block * 25 = 5x5 block

kerneldict

array

build(dq)

Make boolean mask from a data quality array.

Parameters:

dq (np.array of uint32) – 2D data quality array.

Returns:

Grown mask; True indicates a masked pixel, False a normal pixel.

Return type:

np.array of bool

convert_file(file_in, file_mask)

Stand-alone function to make a mask from a file.

The type of output depends on the file_mask extension: * If .asdf is requested, simply writes the boolean array. * If .fits is requested, makes a masked image (HDU0, for display purposes)

and an int8 version (HDU1).

Parameters:

• file_in (str) – The input ASDF file, in L2 format.

• file_mask (str) – The output file.

Return type:

None

PixelMask1