Assemble a coadded image from a set of coadded temporary exposures, being careful to clip & flag areas with potential artifacts. More...
Public Member Functions | |
def | __init__ |
Initialize the task and make the clipDetection subtask. More... | |
def | assemble |
Assemble the coadd for a region. More... | |
def | buildDifferenceImage |
Return an exposure that contains the difference between and unclipped and clipped coadds. More... | |
def | detectClip |
Detect clipped regions on an exposure and set the mask on the individual tempExp masks. More... | |
def | detectClipBig |
Find footprints from individual tempExp footprints for large footprints. More... | |
def | run |
Assemble a coadd from a set of Warps. More... | |
def | getTempExpRefList |
Generate list data references corresponding to warped exposures that lie within the patch to be coadded. More... | |
def | getBackgroundReferenceScaler |
Construct an image scaler for the background reference frame. More... | |
def | prepareInputs |
Prepare the input warps for coaddition by measuring the weight for each warp and the scaling for the photometric zero point. More... | |
def | backgroundMatching |
Perform background matching on the prepared inputs. More... | |
def | assembleMetadata |
Set the metadata for the coadd. More... | |
def | assembleSubregion |
Assemble the coadd for a sub-region. More... | |
def | addBackgroundMatchingMetadata |
Add metadata from the background matching to the coadd. More... | |
def | readBrightObjectMasks |
def | setBrightObjectMasks |
Public Attributes | |
brightObjectBitmask | |
warpType | |
Static Public Attributes | |
ConfigClass = SafeClipAssembleCoaddConfig | |
Assemble a coadded image from a set of coadded temporary exposures, being careful to clip & flag areas with potential artifacts.
Read the documentation for AssembleCoaddTask first since SafeClipAssembleCoaddTask subtasks that task. In AssembleCoaddTask, we compute the coadd as an clipped mean (i.e. we clip outliers). The problem with doing this is that when computing the coadd PSF at a given location, individual visit PSFs from visits with outlier pixels contribute to the coadd PSF and cannot be treated correctly. In this task, we correct for this behavior by creating a new badMaskPlane 'CLIPPED'. We populate this plane on the input coaddTempExps and the final coadd where i. difference imaging suggests that there is an outlier and ii. this outlier appears on only one or two images. Such regions will not contribute to the final coadd. Furthermore, any routine to determine the coadd PSF can now be cognizant of clipped regions. Note that the algorithm implemented by this task is preliminary and works correctly for HSC data. Parameter modifications and or considerable redesigning of the algorithm is likley required for other surveys.
SafeClipAssembleCoaddTask uses a clipDetection subtask and also sub-classes AssembleCoaddTask. You can retarget the clipDetection subtask if you wish.
Initialize the task and make the clipDetection subtask.
Assemble a coadd from a set of Warps. Coadd a set of Warps. Compute weights to be applied to each Warp and find scalings to match the photometric zeropoint to a reference Warp. Optionally, match backgrounds across Warps if the background has not already been removed. Assemble the Warps using assemble. Interpolate over NaNs and optionally write the coadd to disk. Return the coadded exposure.
[in] | dataRef,: | Data reference defining the patch for coaddition and the reference Warp (if config.autoReference=False). Used to access the following data products:
|
[in] | selectDataList[in],: | List of data references to Warps. Data to be coadded will be selected from this list based on overlap with the patch defined by dataRef. |
See SafeClipAssembleCoaddConfig
The command line task interface supports a flag -d
to import debug.py from your PYTHONPATH
; see baseDebug for more about debug.py files. SafeClipAssembleCoaddTask has no debug variables of its own. The clipDetection subtasks may support debug variables. See the documetation for clipDetection for further information.
SafeClipAssembleCoaddTask assembles a set of warped coaddTempExp images into a coadded image. The SafeClipAssembleCoaddTask is invoked by running assembleCoadd.py without the flag '–legacyCoadd'. Usage of assembleCoadd.py expects a data reference to the tract patch and filter to be coadded (specified using '–id = [KEY=VALUE1[^VALUE2[^VALUE3...] [KEY=VALUE1[^VALUE2[^VALUE3...] ...]]') along with a list of coaddTempExps to attempt to coadd (specified using '–selectId [KEY=VALUE1[^VALUE2[^VALUE3...] [KEY=VALUE1[^VALUE2[^VALUE3...] ...]]'). Only the coaddTempExps that cover the specified tract and patch will be coadded. A list of the available optional arguments can be obtained by calling assembleCoadd.py with the –help command line argument:
To demonstrate usage of the SafeClipAssembleCoaddTask in the larger context of multi-band processing, we will generate the HSC-I & -R band coadds from HSC engineering test data provided in the ci_hsc package. To begin, assuming that the lsst stack has been already set up, we must set up the obs_subaru and ci_hsc packages. This defines the environment variable $CI_HSC_DIR and points at the location of the package. The raw HSC data live in the $CI_HSC_DIR/raw directory. To begin assembling the coadds, we must first
We can perform all of these steps by running
This will produce warped coaddTempExps for each visit. To coadd the wraped data, we call assembleCoadd.py as follows:
This will process the HSC-I band data. The results are written in $CI_HSC_DIR/DATA/deepCoadd-results/HSC-I You may also choose to run:
to generate the coadd for the HSC-R band if you are interested in following multiBand Coadd processing as discussed in High-level Overview of Multi-Band Coadd Processing.
Definition at line 1005 of file assembleCoadd.py.
def lsst.pipe.tasks.assembleCoadd.SafeClipAssembleCoaddTask.__init__ | ( | self, | |
args, | |||
kwargs | |||
) |
Initialize the task and make the clipDetection subtask.
Definition at line 1110 of file assembleCoadd.py.
|
inherited |
Add metadata from the background matching to the coadd.
[in] | coaddExposure,: | Coadd |
[in] | tempExpRefList,: | List of data references for temp exps to go into coadd |
[in] | backgroundInfoList,: | List of background info, results from background matching |
Definition at line 736 of file assembleCoadd.py.
def lsst.pipe.tasks.assembleCoadd.SafeClipAssembleCoaddTask.assemble | ( | self, | |
skyInfo, | |||
tempExpRefList, | |||
imageScalerList, | |||
weightList, | |||
bgModelList, | |||
args, | |||
kwargs | |||
) |
Assemble the coadd for a region.
Compute the difference of coadds created with and without outlier rejection to identify coadd pixels that have outlier values in some individual visits. Detect clipped regions on the difference image and mark these regions on the one or two individual coaddTempExps where they occur if there is significant overlap between the clipped region and a source. This leaves us with a set of footprints from the difference image that have been identified as having occured on just one or two individual visits. However, these footprints were generated from a difference image. It is conceivable for a large diffuse source to have become broken up into multiple footprints acrosss the coadd difference in this process. Determine the clipped region from all overlapping footprints from the detected sources in each visit - these are big footprints. Combine the small and big clipped footprints and mark them on a new bad mask plane Generate the coadd using AssembleCoaddTask.assemble without outlier removal. Clipped footprints will no longer make it into the coadd because they are marked in the new bad mask plane.
N.b. *args and **kwargs are passed but ignored in order to match the call signature expected by the parent task.
skyInfo,: | Patch geometry information, from getSkyInfo |
tempExpRefList,: | List of data reference to tempExp |
imageScalerList,: | List of image scalers |
weightList,: | List of weights |
bgModelList,: | List of background models from background matching return coadd exposure |
Definition at line 1118 of file assembleCoadd.py.
|
inherited |
Set the metadata for the coadd.
This basic implementation simply sets the filter from the first input.
[in] | coaddExposure,: | The target image for the coadd |
[in] | tempExpRefList,: | List of data references to tempExp |
[in] | weightList,: | List of weights |
Definition at line 628 of file assembleCoadd.py.
|
inherited |
Assemble the coadd for a sub-region.
For each coaddTempExp, check for (and swap in) an alternative mask if one is passed. If background matching is enabled, add the background and background variance from each coaddTempExp. Remove mask planes listed in config.removeMaskPlanes, Finally, stack the actual exposures using statisticsStack with the statistic specified by statsFlags. Typically, the statsFlag will be one of afwMath.MEAN for a mean-stack or afwMath.MEANCLIP for outlier rejection using an N-sigma clipped mean where N and iterations are specified by statsCtrl. Assign the stacked subregion back to the coadd.
[in] | coaddExposure,: | The target image for the coadd |
[in] | bbox,: | Sub-region to coadd |
[in] | tempExpRefList,: | List of data reference to tempExp |
[in] | imageScalerList,: | List of image scalers |
[in] | weightList,: | List of weights |
[in] | bgInfoList,: | List of background data from background matching |
[in] | altMaskList,: | List of alternate masks to use rather than those stored with tempExp, or None |
[in] | statsFlags,: | afwMath.Property object for statistic for coadd |
[in] | statsCtrl,: | Statistics control object for coadd |
Definition at line 674 of file assembleCoadd.py.
|
inherited |
Perform background matching on the prepared inputs.
Each Warp has a different background level that must be normalized to a reference level before coaddition. If no reference is provided, the background matcher selects one. If the background matching is performed sucessfully, recompute the weight to be applied to the Warp (coaddTempExp) to be consistent with the scaled background.
[in] | inputData,: | Struct from prepareInputs() with tempExpRefList, weightList, imageScalerList |
[in] | refExpDataRef,: | Data reference for background reference Warp, or None |
[in] | refImageScaler,: | Image scaler for background reference Warp, or None |
Definition at line 494 of file assembleCoadd.py.
def lsst.pipe.tasks.assembleCoadd.SafeClipAssembleCoaddTask.buildDifferenceImage | ( | self, | |
skyInfo, | |||
tempExpRefList, | |||
imageScalerList, | |||
weightList, | |||
bgModelList | |||
) |
Return an exposure that contains the difference between and unclipped and clipped coadds.
Generate a difference image between clipped and unclipped coadds. Compute the difference image by subtracting an outlier-clipped coadd from an outlier-unclipped coadd. Return the difference image.
skyInfo,: | Patch geometry information, from getSkyInfo |
tempExpRefList,: | List of data reference to tempExp |
imageScalerList,: | List of image scalers |
weightList,: | List of weights |
bgModelList,: | List of background models from background matching |
Definition at line 1183 of file assembleCoadd.py.
def lsst.pipe.tasks.assembleCoadd.SafeClipAssembleCoaddTask.detectClip | ( | self, | |
exp, | |||
tempExpRefList | |||
) |
Detect clipped regions on an exposure and set the mask on the individual tempExp masks.
Detect footprints in the difference image after smoothing the difference image with a Gaussian kernal. Identify footprints that overlap with one or two input coaddTempExps by comparing the computed overlap fraction to thresholds set in the config. A different threshold is applied depending on the number of overlapping visits (restricted to one or two). If the overlap exceeds the thresholds, the footprint is considered "CLIPPED" and is marked as such on the coaddTempExp. Return a struct with the clipped footprints, the indices of the coaddTempExps that end up overlapping with the clipped footprints and a list of new masks for the coaddTempExps.
[in] | exp,: | Exposure to run detection on |
[in] | tempExpRefList,: | List of data reference to tempExp |
Definition at line 1220 of file assembleCoadd.py.
def lsst.pipe.tasks.assembleCoadd.SafeClipAssembleCoaddTask.detectClipBig | ( | self, | |
tempExpClipList, | |||
clipFootprints, | |||
clipIndices, | |||
maskClipValue, | |||
maskDetValue | |||
) |
Find footprints from individual tempExp footprints for large footprints.
Identify big footprints composed of many sources in the coadd difference that may have originated in a large diffuse source in the coadd. We do this by indentifying all clipped footprints that overlap significantly with each source in all the coaddTempExps.
[in] | tempExpClipList,: | List of tempExp masks with clipping information |
[in] | clipFootprints,: | List of clipped footprints |
[in] | clipIndices,: | List of which entries in tempExpClipList each footprint belongs to |
[in] | maskClipValue,: | Mask value of clipped pixels |
[in] | maskClipValue,: | Mask value of detected pixels |
Definition at line 1316 of file assembleCoadd.py.
|
inherited |
Construct an image scaler for the background reference frame.
Each Warp has a different background level. A reference background level must be chosen before coaddition. If config.autoReference=True, backgroundMatching will pick the reference level and this routine is a no-op and None is returned. Otherwise, use the scaleZeroPoint subtask to compute an imageScaler object for the provided reference image and return it.
[in] | dataRef,: | Data reference for the background reference frame, or None |
Definition at line 403 of file assembleCoadd.py.
|
inherited |
Generate list data references corresponding to warped exposures that lie within the patch to be coadded.
[in] | patchRef,: | Data reference for patch |
[in] | calExpRefList,: | List of data references for input calexps |
Definition at line 386 of file assembleCoadd.py.
|
inherited |
Prepare the input warps for coaddition by measuring the weight for each warp and the scaling for the photometric zero point.
Each Warp has its own photometric zeropoint and background variance. Before coadding these Warps together, compute a scale factor to normalize the photometric zeropoint and compute the weight for each Warp.
[in] | refList,: | List of data references to tempExp |
Definition at line 431 of file assembleCoadd.py.
|
inherited |
Returns None on failure
Definition at line 759 of file assembleCoadd.py.
|
inherited |
Assemble a coadd from a set of Warps.
Coadd a set of Warps. Compute weights to be applied to each Warp and find scalings to match the photometric zeropoint to a reference Warp. Optionally, match backgrounds across Warps if the background has not already been removed. Assemble the Warps using assemble. Interpolate over NaNs and optionally write the coadd to disk. Return the coadded exposure.
[in] | dataRef,: | Data reference defining the patch for coaddition and the reference Warp (if config.autoReference=False). Used to access the following data products:
|
[in] | selectDataList[in],: | List of data references to Warps. Data to be coadded will be selected from this list based on overlap with the patch defined by dataRef. |
Definition at line 320 of file assembleCoadd.py.
|
inherited |
Set the bright object masks exposure: Exposure under consideration dataId: Data identifier dict for patch brightObjectMasks: afwTable of bright objects to mask
Definition at line 767 of file assembleCoadd.py.
|
inherited |
Definition at line 306 of file assembleCoadd.py.
|
static |
Definition at line 1107 of file assembleCoadd.py.
|
inherited |
Definition at line 313 of file assembleCoadd.py.