Injecting Synthetic Sources Into Single-Visit Images¶
Contact author: Jeff Carlin
Last verified to run: 2024-12-02
LSST Science Pipelines version: Weekly 2024_42
Container Size: medium
Targeted learning level: advanced
Description: This tutorial demonstrates a method to inject artificial sources (stars and galaxies) into calexp images using the measured point-spread function of the given calexp image. Confirmation that the synthetic sources were correctly injected into the image is done by running a difference imaging task from the pipelines.
Skills: Use the source_injection tools to inject synthetic sources into images. Create a difference image from a calexp with injected sources.
LSST Data Products: Butler calexp images and corresponding src catalogs, goodSeeingDiff_templateExp images, and injection_catalogs.
Packages: astropy.io, astropy.coordinates, astropy.table, astropy.units, matplotlib, numpy, lsst.afw.display, lsst.afw.table, lsst.daf.base, lsst.daf.butler, lsst.ip.diffim, lsst.source.injection
Credit: Developed by Jeff Carlin in collaboration with Lee Kelvin and the Rubin Community Science Team. Much of the material is based on the documentation for the LSST source_injection package.
Get Support: Find DP0-related documentation and resources at dp0.lsst.io. Questions are welcome as new topics in the Support - Data Preview 0 Category of the Rubin Community Forum. Rubin staff will respond to all questions posted there.
1. Introduction¶
This notebook provides a brief introduction to the source_injection package from the LSST Science Pipelines. The source_injection tools can be used to inject sources into images produced during various stages of Pipelines processing, including visit-level images (e.g., calexps), any dataset with a datasetType of Exposure (e.g., postISRCCD images), and coadd images (e.g., deepCoadds). The main driver that both creates and injects synthetic sources into images is based on Galsim, so that the source_injection tools enable injection of many types of sources. These include a variety of parameterized galaxy models, stars, and postage stamp images.
This notebook will teach some basics of using the source_injection package to insert artificial sources into images. Read the extensive documentation for more details and examples. Furthermore, the Galsim documentation provides much more detail on the types of sources that can be injected.
1.1 Import packages¶
numpy is a commonly used package for scientific computing with arrays in Python.
matplotlib is a library for creating static, animated, and interactive visualizations in Python (see also the matplotlib gallery).
astropy is a package with many common astronomical utilities; here it is used for importing a FITS image, and working with coordinates and tables.
From the lsst package, modules for accessing the butler, image display functions, and the source injection utilities are imported (see pipelines.lsst.io for documentation of the Science Pipelines codebase).
import numpy as np
import matplotlib.pyplot as plt
import os
from astropy.io import fits
import astropy.units as u
from astropy.table import Table, vstack
from astropy.coordinates import SkyCoord
import getpass
import lsst.afw.display as afwDisplay
import lsst.afw.geom as afwGeom
import lsst.afw.image as afwImage
import lsst.afw.math as afwMath
from lsst.daf.butler import Butler
import lsst.geom as geom
from lsst.source.injection import generate_injection_catalog
from lsst.source.injection import VisitInjectConfig, VisitInjectTask
from lsst.ip.diffim.subtractImages import AlardLuptonSubtractTask, AlardLuptonSubtractConfig
1.2 Define functions and parameters¶
Set afwDisplay to use matplotlib, and use colorblind-friendly colors.
afwDisplay.setDefaultBackend('matplotlib')
plt.style.use('tableau-colorblind10')
The following function is a modified version of this code.
def rotate_exposure(exp, n_degrees):
"""Rotate an exposure by nDegrees clockwise.
Parameters
----------
exp : `lsst.afw.image.exposure.Exposure`
The exposure to rotate
n_degrees : `float`
Number of degrees clockwise to rotate by
Returns
-------
rotated_exp : `lsst.afw.image.exposure.Exposure`
A copy of the input exposure, rotated by nDegrees
"""
n_degrees = n_degrees % 360
wcs = exp.getWcs()
warper = afwMath.Warper('lanczos4')
affine_rot_transform = geom.AffineTransform.makeRotation(n_degrees*geom.degrees)
transform_p2top2 = afwGeom.makeTransform(affine_rot_transform)
rotated_wcs = afwGeom.makeModifiedWcs(transform_p2top2, wcs, False)
rotated_exp = warper.warpExposure(rotated_wcs, exp)
return rotated_exp
2. Get an image to inject sources into¶
First, retrieve an image to inject sources into. To do so, instantiate a butler pointing to the DP0.2 dataset, identify calexp images overlapping a particular tract, and retrieve one of those images (plus some of its ancillary data) for later use.
2.1 Instantiate a butler¶
butler_config = 'dp02'
collections = '2.2i/runs/DP0.2'
butler = Butler(butler_config, collections=collections)
2.2 Identify images overlapping a particular tract and select one¶
By default the next cell selects images from tract 3828, but you can change this to any tract number that exists in DP0.2 data. (See the map in Figure 15 of "The DC2 Simulated Sky Survey" overview paper; also visible in the DP0.2 documentation).
Because a calexp image only spans one detector, select a single detector (in this case, number 19, but change this if you like) and find images from the chosen tract that contain that detector.
tract = 3828
where = f"instrument='LSSTCam-imSim' AND skymap='DC2' AND \
tract={tract} AND detector=19 AND band='g'"
calexp_g_DatasetRefs = sorted(list(set(butler.query_datasets('calexp', where=where))))
print(f'Identified {len(calexp_g_DatasetRefs)} calexp DatasetRefs')
Identified 23 calexp DatasetRefs
Use index "5" to select an arbitrary dataId.
Change this index to another integer to select a different image.
dataId_g = calexp_g_DatasetRefs[5].dataId
print(f"{dataId_g = }")
dataId_g = {instrument: 'LSSTCam-imSim', detector: 19, visit: 400440, band: 'g', physical_filter: 'g_sim_1.4', visit_system: 1}
2.2.1 Retrieve the calexp image¶
calexp_g = butler.get('calexp', dataId=dataId_g)
2.2.2 Retrieve additional information about the image¶
To generate synthetic sources to be injected into the image, the coordinates of its bounding box are necessary.
Retrieve the image's WCS and bounding box, and print its central coordinate to the screen:
wcs = calexp_g.getWcs()
bbox = calexp_g.getBBox()
print('bounding box: ', bbox)
boxcen = bbox.getCenter()
cen = wcs.pixelToSky(boxcen)
sc_cen = SkyCoord(ra=cen[0].asDegrees()*u.deg, dec=cen[1].asDegrees()*u.deg)
print(sc_cen)
bounding box: (minimum=(0, 0), maximum=(4071, 3999))
<SkyCoord (ICRS): (ra, dec) in deg
(56.55131893, -36.42591029)>
2.2.3 Figure out how large the image is on the sky¶
Note that above the bounding box is roughly 4000 x 4000 pixels.
Use the bbox.getDimensions and wcs.getPixelScale methods to estimate how large this image is in sky coordinates.
imsize = bbox.getDimensions()[0]*wcs.getPixelScale().asDegrees()
print('Size of calexp in degrees: ', imsize)
Size of calexp in degrees: 0.22586221347938068
3. Make a catalog of synthetic sources¶
Having retrieved a calexp image to inject into, now set up a simple synthetic source catalog.
3.1 Make a catalog of galaxies and stars¶
The source_injection package provides tools to create catalogs of synthetic sources. Here, use the generate_injection_catalog function to create the sources to inject.
Note that the "inject_size" is selected to be 0.1 degrees, or slightly smaller than the size of the image as determined above (inject_size is a radius, so it equals 0.2/2 degrees).
inject_size = 0.2/2
3.1.1 Make a catalog of galaxies to inject¶
The simplest form of a galaxy in Galsim is parameterized by a Sersic model: $I(r) = I_e~{\rm exp}\{-b_n [(\frac{r}{r_e})^{1/n}-1]\}$, which defines the shape of the galaxy's light profile as a function of radius (r) in terms of the "Sersic index" (n) and the "half-light radius" ($r_e$). (Note that $n = 1$ corresponds to an exponential profile.)
The above equation results in a circular galaxy. To create elongated (elliptical) Sersic profiles, additionally specify the minor-to-major axis ratio (q) with a rotation angle (beta).
generate_injection_catalog will automatically generate sources with all possible permutations of the parameters you provide. For example, the cell below specifies "number=2" to create 2 synthetic galaxies, but then specifies a single magnitude (mag), 3 values of Sersic index (n), 3 values of axis ratio (q), 2 values of beta, and a single value for half_light_radius. This should result in $2*1*3*3*2*1 = 36$ different combinations of those parameters.
Finally, provide minimum and maximum RA and Dec coordinates in degrees. In this case, generate_injection_catalog will randomly select positions within those limits.
my_injection_catalog_galaxies = generate_injection_catalog(
ra_lim=[sc_cen.ra.value-inject_size, sc_cen.ra.value+inject_size],
dec_lim=[sc_cen.dec.value-inject_size, sc_cen.dec.value+inject_size],
number=2,
seed='3210',
source_type="Sersic",
mag=[15.0],
n=[1, 2, 4],
q=[0.9, 0.5, 0.1],
beta=[31.0, 144.0],
half_light_radius=[15.0],
)
lsst.source.injection.utils._generate_injection_catalog INFO: Generated an injection catalog containing 36 sources: 18 combinations repeated 2 times.
3.1.2 Make a catalog of stars to inject¶
Using a similar method, create a catalog of stars to be injected. Specify 5 values of magnitude, with 7 instances, which will result in 35 stars.
my_injection_catalog_stars = generate_injection_catalog(
ra_lim=[sc_cen.ra.value-inject_size, sc_cen.ra.value+inject_size],
dec_lim=[sc_cen.dec.value-inject_size, sc_cen.dec.value+inject_size],
number=7,
seed='432',
source_type="Star",
mag=[16.0, 17.0, 18.0, 19.0, 20.0],
)
lsst.source.injection.utils._generate_injection_catalog INFO: Generated an injection catalog containing 35 sources: 5 combinations repeated 7 times.
inject_cat = vstack([my_injection_catalog_galaxies, my_injection_catalog_stars])
inject_cat
| injection_id | ra | dec | source_type | mag | n | q | beta | half_light_radius |
|---|---|---|---|---|---|---|---|---|
| int64 | float64 | float64 | str6 | float64 | int64 | float64 | float64 | float64 |
| 0 | 56.64161200441393 | -36.459276165910296 | Sersic | 15.0 | 1 | 0.9 | 31.0 | 15.0 |
| 1 | 56.52286200441393 | -36.33581937578684 | Sersic | 15.0 | 1 | 0.9 | 31.0 | 15.0 |
| 10 | 56.46036200441393 | -36.362979869614 | Sersic | 15.0 | 1 | 0.9 | 144.0 | 15.0 |
| 11 | 56.45411200441393 | -36.51112801776215 | Sersic | 15.0 | 1 | 0.9 | 144.0 | 15.0 |
| 20 | 56.597862004413926 | -36.46915270912017 | Sersic | 15.0 | 1 | 0.5 | 31.0 | 15.0 |
| 21 | 56.54161200441393 | -36.444461351095484 | Sersic | 15.0 | 1 | 0.5 | 31.0 | 15.0 |
| 30 | 56.504112004413926 | -36.355572462206595 | Sersic | 15.0 | 1 | 0.5 | 144.0 | 15.0 |
| 31 | 56.57286200441393 | -36.35804159800906 | Sersic | 15.0 | 1 | 0.5 | 144.0 | 15.0 |
| 40 | 56.466612004413925 | -36.41483172146585 | Sersic | 15.0 | 1 | 0.1 | 31.0 | 15.0 |
| ... | ... | ... | ... | ... | ... | ... | ... | ... |
| 34 | 56.62302474605745 | -36.490940779804866 | Star | 19.0 | -- | -- | -- | -- |
| 35 | 56.541774746057456 | -36.38723707610116 | Star | 19.0 | -- | -- | -- | -- |
| 36 | 56.566774746057455 | -36.48353337239746 | Star | 19.0 | -- | -- | -- | -- |
| 40 | 56.573024746057456 | -36.37982966869375 | Star | 20.0 | -- | -- | -- | -- |
| 41 | 56.635524746057456 | -36.409459298323384 | Star | 20.0 | -- | -- | -- | -- |
| 42 | 56.604274746057456 | -36.47118769338511 | Star | 20.0 | -- | -- | -- | -- |
| 43 | 56.52302474605745 | -36.357607446471526 | Star | 20.0 | -- | -- | -- | -- |
| 44 | 56.56052474605745 | -36.45390374276783 | Star | 20.0 | -- | -- | -- | -- |
| 45 | 56.64177474605746 | -36.52057040943449 | Star | 20.0 | -- | -- | -- | -- |
| 46 | 56.46052474605745 | -36.49834818721227 | Star | 20.0 | -- | -- | -- | -- |
3.2 Add a postage stamp of an image to the injection catalog¶
In addition to parameterized sources of many types, a postage-stamp image (for example, an output image from a simulation) can also be injected. This can be any image; the only requirement is that it must be in FITS format and have a valid WCS (see a fun example here). This notebook demonstrates injection of an SDSS g-band image of the galaxy PGC 038749 downloaded from the NASA/IPAC Extragalactic Database (NED).
Load the image, then display it to see what it depicts:
stamp_img_hdu = fits.open('data/PGC_038749_I_g_bbl2011.fits')
fig = plt.figure()
plt.subplot()
im = plt.imshow(stamp_img_hdu[0].data, cmap='gray', vmin=-20.0, vmax=100, origin='lower')
plt.xlabel('x')
plt.ylabel('y')
plt.show()
Figure 1: The SDSS g-band image of PGC 038749 displayed in grayscale, with axes in pixel coordinates.
3.2.1 Create an entry in the injection catalog for the postage stamp image¶
Add a postage stamp entry into the injection catalog, using vstack to add an astropy Table containing a single row. Specify its position, magnitude, and "source_type = Stamp," in addition to the filename of the image to inject. (It is OK to leave unnecessary columns empty.)
my_injection_catalog_stamp = Table(
{
'injection_id': [9999],
'ra': [56.622],
'dec': [-36.488],
'source_type': ['Stamp'],
'mag': [14.8],
'stamp': ['data/PGC_038749_I_g_bbl2011.fits'],
}
)
inject_cat = vstack([inject_cat, my_injection_catalog_stamp])
inject_cat
| injection_id | ra | dec | source_type | mag | n | q | beta | half_light_radius | stamp |
|---|---|---|---|---|---|---|---|---|---|
| int64 | float64 | float64 | str6 | float64 | int64 | float64 | float64 | float64 | str32 |
| 0 | 56.64161200441393 | -36.459276165910296 | Sersic | 15.0 | 1 | 0.9 | 31.0 | 15.0 | -- |
| 1 | 56.52286200441393 | -36.33581937578684 | Sersic | 15.0 | 1 | 0.9 | 31.0 | 15.0 | -- |
| 10 | 56.46036200441393 | -36.362979869614 | Sersic | 15.0 | 1 | 0.9 | 144.0 | 15.0 | -- |
| 11 | 56.45411200441393 | -36.51112801776215 | Sersic | 15.0 | 1 | 0.9 | 144.0 | 15.0 | -- |
| 20 | 56.597862004413926 | -36.46915270912017 | Sersic | 15.0 | 1 | 0.5 | 31.0 | 15.0 | -- |
| 21 | 56.54161200441393 | -36.444461351095484 | Sersic | 15.0 | 1 | 0.5 | 31.0 | 15.0 | -- |
| 30 | 56.504112004413926 | -36.355572462206595 | Sersic | 15.0 | 1 | 0.5 | 144.0 | 15.0 | -- |
| 31 | 56.57286200441393 | -36.35804159800906 | Sersic | 15.0 | 1 | 0.5 | 144.0 | 15.0 | -- |
| 40 | 56.466612004413925 | -36.41483172146585 | Sersic | 15.0 | 1 | 0.1 | 31.0 | 15.0 | -- |
| ... | ... | ... | ... | ... | ... | ... | ... | ... | ... |
| 35 | 56.541774746057456 | -36.38723707610116 | Star | 19.0 | -- | -- | -- | -- | -- |
| 36 | 56.566774746057455 | -36.48353337239746 | Star | 19.0 | -- | -- | -- | -- | -- |
| 40 | 56.573024746057456 | -36.37982966869375 | Star | 20.0 | -- | -- | -- | -- | -- |
| 41 | 56.635524746057456 | -36.409459298323384 | Star | 20.0 | -- | -- | -- | -- | -- |
| 42 | 56.604274746057456 | -36.47118769338511 | Star | 20.0 | -- | -- | -- | -- | -- |
| 43 | 56.52302474605745 | -36.357607446471526 | Star | 20.0 | -- | -- | -- | -- | -- |
| 44 | 56.56052474605745 | -36.45390374276783 | Star | 20.0 | -- | -- | -- | -- | -- |
| 45 | 56.64177474605746 | -36.52057040943449 | Star | 20.0 | -- | -- | -- | -- | -- |
| 46 | 56.46052474605745 | -36.49834818721227 | Star | 20.0 | -- | -- | -- | -- | -- |
| 9999 | 56.622 | -36.488 | Stamp | 14.8 | -- | -- | -- | -- | data/PGC_038749_I_g_bbl2011.fits |
3.2.2 Create a rotated version of the stamp image, and add an entry for it into the injection catalog¶
To apply a rotation to the postage stamp image before injecting it into the calexp, use the rotateExposure function created above.
The process is to read the image, apply a rotation to it, write it out as a new image, and then follow the same steps as the previous section to add it to the injection catalog.
First, read the postage stamp image using the afwImage package. It's OK to ignore the warning about an expected extension type that will pop up.
stamp_img_orig = afwImage.ExposureF.readFits('data/PGC_038749_I_g_bbl2011.fits')
lsst.afw.image.MaskedImageFitsReader WARNING: Expected extension type not found: IMAGE
Apply a rotation to the image using the rotateExposure function defined above. This will apply a rotation to the WCS, and then "warp" the image (i.e., resample it) onto the new, rotated WCS.
rotation_angle = 57.0
stamp_img_rotated = rotate_exposure(stamp_img_orig, rotation_angle)
Confirm that the image has been rotated by displaying them side-by-side:
fig, ax = plt.subplots(1, 2, figsize=(6, 4), dpi=150)
plt.sca(ax[0])
display0 = afwDisplay.Display(frame=fig)
display0.scale('linear', min=-20, max=150)
display0.mtv(stamp_img_orig.image)
plt.title('original stamp image')
plt.sca(ax[1])
display1 = afwDisplay.Display(frame=fig)
display1.scale('linear', min=-20, max=150)
display1.mtv(stamp_img_rotated.image)
plt.title('rotated stamp image')
plt.tight_layout()
plt.show()
Figure 2: At left, the same image as shown in Figure 1. At right, the image rotated by 57 degrees clockwise (west of north), as if the telescope had a rotation angle of 57 degrees east of north.
When the image is rotated, the "empty" pixels are given NaN values. Replace those with zeros (if you leave them as NaNs, the image values will not be read correctly when reading in the resulting image).
stamp_img_rotated.image.array[np.where(np.isnan(stamp_img_rotated.image.array))] = 0.0
Write the rotated stamp to an output file in the home directory.
home_directory = '/home/' + getpass.getuser() + '/'
stamp_img_rotated.writeFits(home_directory+'stamp_rotated.fits')
Add an entry to the injection catalog for the rotated stamp.
my_injection_catalog_rotated_stamp = Table(
{
'injection_id': [99999],
'ra': [56.642],
'dec': [-36.488],
'source_type': ['Stamp'],
'mag': [14.8],
'stamp': [home_directory+'stamp_rotated.fits'],
}
)
inject_cat = vstack([inject_cat, my_injection_catalog_rotated_stamp])
4. Inject sources into a calexp image¶
4.1 Use source_injection tools to inject sources¶
In Section 3, a catalog specifying the synthetic sources to inject was created.
Now run the task from source_injection that injects sources into the calexp image that was retrieved earlier.
First, extract the point spread function (PSF), photometric calibration object, and the WCS (World Coordinate System) object associated with the calexp image. These will be passed to the injection task so that sources can be injected using the properties of the image itself.
psf = calexp_g.getPsf()
photo_calib = calexp_g.getPhotoCalib()
wcs = calexp_g.getWcs()
4.2 Instantiate the injection classes¶
First, instantiate the VisitInjectConfig class. The VisitInjectConfig class is where configuration of the injection task occurs, allowing for modifications to be made to how the task operates.
Then instantiate the VisitInjectTask, using inject_config as the configuration argument.
NOTE: For injections into other dataset types, use the appropriate option from the following list:
from lsst.source.injection import CoaddInjectConfig, CoaddInjectTask
from lsst.source.injection import ExposureInjectConfig, ExposureInjectTask
from lsst.source.injection import VisitInjectConfig, VisitInjectTask
Both ExposureInject and VisitInject accept dimensions of ("instrument", "visit", "detector"). They differ in that ExposureInject takes a postISRCCD exposure (i.e., a raw image that has had "instrument signature removal," or ISR, applied, and no further processing) as an input, while VisitInject expects to operate on a calexp (an image that has been astrometrically and photometrically calibrated to an external source).
inject_config = VisitInjectConfig()
inject_task = VisitInjectTask(config=inject_config)
4.3 Inject sources into an image¶
Next, execute the run method of the inject_task.
The run method needs the following inputs:
- the input injection catalog
- the input
calexp - the PSF of the input exposure
- the WCS information
- the photometric calibration information
The PSF, WCS, and photo_calib inputs were already retrieved directly from the calexp they are associated with.
The inject task provides two outputs:
- the output exposure with sources injected
- the output source injection catalog
The output source injection catalog is identical to the input, but with two additional columns (x and y) denoting the pixel coordinates of these sources. Note that this catalog is NOT the science catalog containing the full suite of LSST Science Pipelines outputs. To get that, this source injected image will need to be processed by additional Science Pipelines tasks.
Note: it is best to use a clone of the input
calexp. This is because thecalexpthat is input for source injection will be edited in-place. Inputting a clone makes it possible to compare the output image to the originalcalexplater in this notebook.
4.3.1 Run the source injection task¶
Run the source injection task and extract the "output_exposure" and "output_catalog":
injected_output = inject_task.run(
injection_catalogs=inject_cat,
input_exposure=calexp_g.clone(),
psf=psf,
photo_calib=photo_calib,
wcs=wcs,
)
injected_exposure = injected_output.output_exposure
injected_catalog = injected_output.output_catalog
lsst.visitInjectTask INFO: Retrieved 73 injection sources from 73 HTM trixels.
lsst.visitInjectTask INFO: Identified 1 injection source with a centroid outside the padded image bounding box.
lsst.visitInjectTask INFO: Catalog cleaning removed 1 of 73 sources; 72 remaining for catalog checking.
lsst.visitInjectTask INFO: Catalog checking flagged 0 of 72 sources; 72 remaining for source generation.
lsst.visitInjectTask INFO: Adding INJECTED and INJECTED_CORE mask planes to the exposure.
lsst.visitInjectTask INFO: Generating 72 injection sources consisting of 3 unique types: Sersic(36), DeltaFunction(34), Stamp(2).
lsst.visitInjectTask WARNING: Clipping draw size for object at (56.5822370044, -36.4469304869) from 11850 to 10000 pixels.
lsst.visitInjectTask WARNING: Clipping draw size for object at (56.5291120044, -36.4666835733) from 11848 to 10000 pixels.
lsst.visitInjectTask WARNING: Clipping draw size for object at (56.5603620044, -36.4296465363) from 11850 to 10000 pixels.
lsst.visitInjectTask WARNING: Clipping draw size for object at (56.4978620044, -36.4518687585) from 11848 to 10000 pixels.
lsst.visitInjectTask INFO: Injected 64 of 72 potential sources. 8 sources flagged and skipped: NO_OVERLAP(1), FFT_SIZE_ERROR(7).
4.3.2 Compare the images before and after injection¶
Plot the images side-by-side to confirm that the source injection task successfully added sources to the image.
Note that the injected image will have different flux values than the original, which would cause the image scaling of the two images to be slightly different by default (to confirm this, uncomment the "display0.scale('linear', 'zscale')" lines below and comment out the lines below them that explicitly set the min/max values). For a direct comparison, explicitly set the minimum and maximum pixel values of the color scale.
Additionally, to zoom in on the injected postage stamp to see how it looks in the image, uncomment the line below where it says "To zoom on the PGC 038749 stamp:".
plot_injected_calexp = injected_exposure.clone()
fig, ax = plt.subplots(1, 2, figsize=(10, 6), dpi=150)
plt.sca(ax[0])
display0 = afwDisplay.Display(frame=fig)
# display0.scale('linear', 'zscale')
display0.scale('linear', min=-20, max=150)
display0.mtv(calexp_g.image)
plt.title('calexp image')
plt.sca(ax[1])
display1 = afwDisplay.Display(frame=fig)
# display1.scale('linear', 'zscale')
display1.scale('linear', min=-20, max=150)
display1.mtv(plot_injected_calexp.image)
# To zoom on the PGC 038749 stamp:
# display1.mtv(plot_injected_calexp.image[3300:3800, 1500:2000])
plt.title('injected_calexp image')
plt.tight_layout()
plt.show()
Figure 3: The
calexpimage displayed in grayscale (with scale bar at right and axes in pixels) before (left) and after (right) synthetic source injection.
4.4 Create a difference image to see the injected sources¶
One reason for injecting sources into a calexp image might be to test whether they would be recovered in a difference image. This section demonstrates the use of an image differencing task from the LSST Science Pipelines to create a difference image.
4.4.1 Load the template image¶
Template images are created as coadds of the images with the best seeing, and are called goodSeeingDiff_templateExp. Load the appropriate template corresponding to the calexp used throughout this notebook (note the use of the same dataId used to load the calexp).
difftemp = 'goodSeeingDiff_templateExp'
templateExposure = butler.get(difftemp, dataId=dataId_g.required)
4.4.2 Initialize and run the image subtraction task¶
The first two lines below load the default configuration for the AlardLuptonSubtract task, and then initialize the task with that configuration. The task requires (1) a template exposure, (2) the science exposure, and (3) the catalog of sources from the science exposure.
Load the source catalog (src) and run the task. Override the default configuration to point the "sourceSelector" to "base_ClassificationExtendedness_value."
As of w_2024_38, the source selector within AlardLuptonSubtractTask has been changed. In the new selector, the default value that is used does not exist in DP0.2 catalogs, so we need to override the configured value to use base_ClassificationExtendedness_value instead. However, that override will fail on all pipeline versions prior to w_38, so we wrap it in a try:except block so that this cell will run with any pipelines version.
config = AlardLuptonSubtractConfig()
try:
config.sourceSelector.value.unresolved.name = 'base_ClassificationExtendedness_value'
except:
pass
alTask = AlardLuptonSubtractTask(config=config)
scienceExposure = injected_exposure
sources = butler.get('src', dataId=dataId_g)
result = alTask.run(templateExposure, scienceExposure, sources)
lsst.alardLuptonSubtract INFO: template has 16286688 good pixels (100.0%)
lsst.alardLuptonSubtract.scaleVariance INFO: Renormalizing variance by 1.151930
lsst.alardLuptonSubtract.scaleVariance INFO: Renormalizing variance by 1.024500
lsst.alardLuptonSubtract INFO: Template variance scaling factor: 1.15
lsst.alardLuptonSubtract INFO: Science variance scaling factor: 1.02
lsst.alardLuptonSubtract INFO: Science PSF FWHM: 3.834993 pixels
lsst.alardLuptonSubtract INFO: Template PSF FWHM: 3.156140 pixels
lsst.alardLuptonSubtract INFO: `convolveTemplate` is set: convolving template image.
lsst.alardLuptonSubtract.sourceSelector INFO: Selected 193/2968 sources
lsst.alardLuptonSubtract INFO: Rejecting 0 candidate sources: an excluded template mask plane is set.
lsst.alardLuptonSubtract INFO: 193/2968=6.5% of sources selected for PSF matching from the input catalog
lsst.alardLuptonSubtract.makeKernel INFO: Reference psf fwhm is the greater, normal convolution mode
lsst.alardLuptonSubtract.makeKernel INFO: Selected 191 / 193 sources as kernel candidates.
lsst.alardLuptonSubtract.makeKernel INFO: Reference psf fwhm is the greater, normal convolution mode
lsst.alardLuptonSubtract.makeKernel INFO: Final spatial kernel sum 137.890
lsst.alardLuptonSubtract.makeKernel INFO: Spatial model condition number 2.783e+07
lsst.alardLuptonSubtract.makeKernel INFO: Doing stats of kernel candidates used in the spatial fit.
lsst.alardLuptonSubtract.makeKernel INFO: 191 candidates total, 122 rejected, 69 used
lsst.alardLuptonSubtract.makeKernel INFO: Spatial kernel model well constrained; 69 candidates, 3 terms, 27 bases
lsst.alardLuptonSubtract.makeKernel INFO: Spatial background model appears well constrained; 69 candidates, 6 terms
lsst.alardLuptonSubtract INFO: Decorrelating image difference.
lsst.alardLuptonSubtract.decorrelate INFO: Using matching kernel computed at (2036, 2000)
lsst.alardLuptonSubtract.decorrelate INFO: Original variance plane means. Science:7.39610e+02, warped template:2.15310e-03)
lsst.alardLuptonSubtract.decorrelate INFO: Decorrelation after template image convolution
lsst.alardLuptonSubtract.decorrelate INFO: Variance plane mean of uncorrected diffim: 743.159544
lsst.alardLuptonSubtract.decorrelate INFO: Decorrelation of difference image
lsst.alardLuptonSubtract.decorrelate INFO: Common frequency space shape (4040, 4112)
lsst.alardLuptonSubtract.decorrelate INFO: Variance plane mean of corrected diffim: 7.80410e+02
4.4.3 Compare the original and the difference image¶
Plot the original calexp, the injected image, and the difference image side by side.
Set xmin, xmax, ymin, and ymax to zoom in on an arbitrary section of the image.
Change the values to zoom in on a different region.
xmin, xmax = 500, 2000
ymin, ymax = 1000, 2500
fig, ax = plt.subplots(1, 3, figsize=(14, 11))
plt.sca(ax[0]) # set the first axis as current
display1 = afwDisplay.Display(frame=fig)
display1.scale('linear', 'zscale')
display1.mtv(calexp_g.image)
plt.title('coadd image (template)')
plt.xlim(xmin, xmax)
plt.ylim(ymin, ymax)
plt.sca(ax[1]) # set the second axis as current
display2 = afwDisplay.Display(frame=fig)
display2.scale('linear', 'zscale')
display2.mtv(plot_injected_calexp.image)
plt.title('calexp with synthetic sources')
plt.xlim(xmin, xmax)
plt.ylim(ymin, ymax)
plt.sca(ax[2]) # set the third axis as current
display3 = afwDisplay.Display(frame=fig)
display3.scale('linear', 'zscale')
display3.mtv(result.difference.image)
plt.xlim(xmin, xmax)
plt.ylim(ymin, ymax)
plt.title('diffim')
plt.tight_layout()
plt.show()
Figure 4: At left, "zoomed-in" cutout of the
deepCoaddimage displayed in grayscale (with scale bar at right and axes in pixels). In the center, the cutout of thecalexpwith synthetic sources injected (as in the right panel of Figure 3). At right, the difference image which results from subtracting thecoaddfrom thecalexp.
5. Exercises for the learner¶
Further explorations could include:
- Injecting sources into coadd images.
- Exploring other types of objects that can be injected into images (i.e., other Galsim parameters for more complicated sources).
- Injecting variable objects into multiple calexp images and testing their recoverability.
- Running detection and measurement tasks from the LSST Science Pipelines to test the recovery and measurement accuracy of injected sources (e.g., following DP0.2 tutorial Notebook 05: Introduction to Source Detection).