romanimpreprocess.from_sim.sim_to_isim

Functions to convert external simulated images to Roman L1/L2-like format.

This works entirely at the single exposure level. Some parts wrap romanisim.

Functions

hdu_sip_hflip: Flips an SCA in the 3n row in Detector coordinates to align with Science coordinates.
hdu_sip_vflip: Flips an SCA in the 3n+1 or 3n+2 rows in Detector coordinates to align with Science coordinates.
make_l1_fullcal: Makes an L1 image using an OpenUniverse input and the calibration data. (Merges with romanisim routines.)
noise_1f_frame: Generates 1/f noise.
fill_in_refdata_and_1f: Fills in reference pixels and reference output, as well as 1/f noise.
simpletest: Quick look tool for Level 1 to Level 2 conversion (not for production).
runconfig: Configuration-driven function to convert a simulation to Level 1 format.

Classes

Image2D: 2D image (may be from simulation).
Image2D_from_L1: 2D image from Level 1 data (useful for ‘shortcut’ workflow, for most applications use L1_to_L2).

Attributes

config

Classes

`Image2D`	2D image, along with WCS and sky information.
`Image2D_from_L1`	Similar to Image2D, but constructed from L1 data file.

Functions

`hdu_sip_hflip`(data, header)	Horizontal flip of SCA and WCS. Assumes SIP convention.
`hdu_sip_vflip`(data, header)	Vertical flip of SCA and WCS. Assumes SIP convention.
`make_l1_fullcal`(counts, read_pattern, caldir[, rng, ...])	Make an L1 image with the full calibration information.
`noise_1f_frame`(rng)	Generates a 4096x128 block of 1/f noise.
`fill_in_refdata_and_1f`(im, caldir, rng, tij[, ...])	Fills in the reference pixel data in an image, and adds 1/f noise.
`simpletest`()	This is a simple script to convert Roman to L1/L2.
`run_config`(config)	This allows the L1 image construction to be called as a Python function instead of a

Module Contents

hdu_sip_hflip(data, header)[source]

Horizontal flip of SCA and WCS. Assumes SIP convention.

This function operates on the data and WCS header in place.

Parameters:

data (np.array) – 2D image of an SCA.
header (astropy.io.fits.header.Header) – Header containing the WCS.

Return type:

None

See also

hdu_sip_hflip: Similar, but for horizontal flip instead.

make_l1_fullcal(counts, read_pattern, caldir, rng=None, persistence=None, tstart=None)[source]

Make an L1 image with the full calibration information.

This carries out similar steps to romanisim.l1.make_l1, but provides us a bit more control over the settings.

Parameters:

counts (galsim.Image) – Contains mean number electrons per pixel per exposure.
read_pattern (list of list of int) – MultiAccum table.
caldir (dict) – Dictionary of the reference files.
rng (galsim.BaseDeviate, optional) – The random number generator.
persistence (romanisim.persistence.Persistence, optional) – Persistence object, not used yet.
tstart (float, optional) – Start time to feed to romanisim.

Returns:

l1 (np.array) – 3D image array
dq (np.array) – 3D quality array

noise_1f_frame(rng)[source]

Generates a 4096x128 block of 1/f noise.

The frame is normalized to variance of 1 per logarithmic range in frequency, i.e., S(f) = 1/f, where Var X = int_0^infty S(f) df.

Parameters:: rng (galsim.BaseDeviate) – The random number generator.
Returns:: Shape (4096,128).
Return type:: np.array of float

fill_in_refdata_and_1f(im, caldir, rng, tij, fill_in_banding=True, amp33=None)[source]

Fills in the reference pixel data in an image, and adds 1/f noise.

If amp33 is provided, then also tries to fill in the reference output (if the calibration reference files have that information).

Noise is added in-place to im and (if provided) amp33, keeping the same data type.

Parameters:

im (np.array) – The image data cube. Shape (ngrp, ny, nx).
caldir (dict) – The dictionary of calibration reference files.
rng (galsim.BaseDeviate) – The random number generator.
tij (list of list of float) – Time stamps of the reads in seconds.
fill_in_banding (bool, optional) – Whether to gennerate 1/f noise.
amp33 (np.array, optional) – If provided, array to put the reference output.

Return type:

None

class Image2D(intype, **kwargs)[source]

2D image, along with WCS and sky information.

It can be constructed from simulations or (ultimately) from Roman data.

Parameters:: intype (str) – Input type (e.g., “anlsim”)

image

A 2D image. Units are e/p/s if generated from a simulation in e.

Type:: np.array of flat

galsimwcs

The world coordinate system for this image.

Type:: galsim.wcs.CelestialWCS

header

The world coordinate system for this image in FITS WCS format.

Type:: astropy.io.fits.header.Header

date

The observation date (ISO 8601 string).

Type:: str

filter

The observation filter (4 characters, e.g., R062).

Type:: str

idsca

The observation ID and SCA.

Type:: (int, int)

ra_

Right ascension (in degrees) of the WFI.

Type:: float

dec_

Declination (in degrees) of the WFI.

Type:: float

pa_

Position angle (in degrees) of the WFI.

Type:: float

refdata

Reference data locations.

Type:: dict

af

Level 1 data

Type:: asdf.AsdfFile

af2

Level 2 data

Type:: asdf.AsdfFile

__init__()[source]: Constructor

init_anlsim()[source]: Constructor from Open Universe simulation file.

simulate()[source]: Simulates the ramps, including L1 and L2 data.

L1_write_to()[source]: Write simulated L1 data file (ASDF)

L2_write_to()[source]: Write simulated L2 data file (ASDF)

Notes

The legal intype strings are:

anlsim : The Open Universe 2024 simulation “truth” (or equivalent)

init_anlsim(fname, flip=True)[source]

Constructor from Open Universe 2024-type simulation.

Parameters:

fname (str) – The input file name.
flip (bool, optional) – If True, then flips from SCA native coordinates to science-aligned (SOC-like product).

Return type:

None

simulate(use_read_pattern, caldir=None, config={}, seed=43)[source]

Performs Level 1 & 2 simulations.

This is based on the romanisim.image.simulate function, but some functionality has been changed to be useful for this class.

Parameters:

use_read_pattern (list of list of int) – The MultiAccum table.
caldir (dict, optional) – Dictionary of where the calibration files are located. (Otherwise uses internal defaults, only good for testing.)
config (dict, optional) – Configuration file (usually expanded from YAML).
seed (int, optional) – Random number seed.

Return type:

None

L1_write_to(filename)[source]

Writes L1 data to a file if available.

Parameters:: filename (str) – Where to write the file (should end with .asdf).
Returns:: True if successful, False if not written.
Return type:: bool

L2_write_to(filename)[source]

Writes L2 data to a file if available.

Parameters:: filename (str) – Where to write the file (should end with .asdf).
Returns:: True if successful, False if not written.
Return type:: bool

class Image2D_from_L1(infile, refdata, thewcs, verbose_err=True)[source]

Bases: Image2D

Similar to Image2D, but constructed from L1 data file.

with Image2D_from_L1(infile, refdata, thewcs) as L1:: …

__enter__()[source]

__exit__(exception_type, exception_value, exception_traceback)[source]

af[source]

refdata[source]

thewcs[source]

psuedocalibrate()[source]

Generates a simple calibrated (L2) image.

This doesn’t use romancal, but can be useful as a pass-through function.

simpletest()[source]: This is a simple script to convert Roman to L1/L2. For internal testing only, not production.

run_config(config)[source]

This allows the L1 image construction to be called as a Python function instead of a stand-alone code.

Parameters:: config (dict) – Configuration file (usually a dictionary unpacked from YAML).
Return type:: None

config[source]