fgcmBuildStarsTable.py

Go to the documentation of this file.

# See COPYRIGHT file at the top of the source tree.
#
# This file is part of fgcmcal.
#
# Developed for the LSST Data Management System.
# This product includes software developed by the LSST Project
# (https://www.lsst.org).
# See the COPYRIGHT file at the top-level directory of this distribution
# for details of code ownership.
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program.  If not, see <https://www.gnu.org/licenses/>.
"""Build star observations for input to FGCM using sourceTable_visit.
 
This task finds all the visits and sourceTable_visits in a repository (or a
subset based on command line parameters) and extracts all the potential
calibration stars for input into fgcm.  This task additionally uses fgcm to
match star observations into unique stars, and performs as much cleaning of the
input catalog as possible.
"""
 
import time
 
import numpy as np
import collections
 
import lsst.daf.persistence as dafPersist
import lsst.pex.config as pexConfig
import lsst.pipe.base as pipeBase
from lsst.pipe.base import connectionTypes
import lsst.afw.table as afwTable
from lsst.meas.algorithms import ReferenceObjectLoader
 
from .fgcmBuildStarsBase import FgcmBuildStarsConfigBase, FgcmBuildStarsRunner, FgcmBuildStarsBaseTask
from .utilities import computeApproxPixelAreaFields, computeApertureRadiusFromDataRef
from .utilities import lookupStaticCalibrations
 
__all__ = ['FgcmBuildStarsTableConfig', 'FgcmBuildStarsTableTask']
 
 
class FgcmBuildStarsTableConnections(pipeBase.PipelineTaskConnections,
                                     dimensions=("instrument",),
                                     defaultTemplates={}):
    camera = connectionTypes.PrerequisiteInput(
        doc="Camera instrument",
        name="camera",
        storageClass="Camera",
        dimensions=("instrument",),
        lookupFunction=lookupStaticCalibrations,
        isCalibration=True,
    )
 
    fgcmLookUpTable = connectionTypes.PrerequisiteInput(
        doc=("Atmosphere + instrument look-up-table for FGCM throughput and "
             "chromatic corrections."),
        name="fgcmLookUpTable",
        storageClass="Catalog",
        dimensions=("instrument",),
        deferLoad=True,
    )
 
    sourceSchema = connectionTypes.PrerequisiteInput(
        doc="Schema for source catalogs",
        name="src_schema",
        storageClass="SourceCatalog",
        deferLoad=True,
    )
 
    refCat = connectionTypes.PrerequisiteInput(
        doc="Reference catalog to use for photometric calibration",
        name="cal_ref_cat",
        storageClass="SimpleCatalog",
        dimensions=("skypix",),
        deferLoad=True,
        multiple=True,
    )
 
    sourceTable_visit = connectionTypes.Input(
        doc="Source table in parquet format, per visit",
        name="sourceTable_visit",
        storageClass="DataFrame",
        dimensions=("instrument", "visit"),
        deferLoad=True,
        multiple=True,
    )
 
    visitSummary = connectionTypes.Input(
        doc=("Per-visit consolidated exposure metadata.  These catalogs use "
             "detector id for the id and must be sorted for fast lookups of a "
             "detector."),
        name="visitSummary",
        storageClass="ExposureCatalog",
        dimensions=("instrument", "visit"),
        deferLoad=True,
        multiple=True,
    )
 
    background = connectionTypes.Input(
        doc="Calexp background model",
        name="calexpBackground",
        storageClass="Background",
        dimensions=("instrument", "visit", "detector"),
        deferLoad=True,
        multiple=True,
    )
 
    fgcmVisitCatalog = connectionTypes.Output(
        doc="Catalog of visit information for fgcm",
        name="fgcmVisitCatalog",
        storageClass="Catalog",
        dimensions=("instrument",),
    )
 
    fgcmStarObservations = connectionTypes.Output(
        doc="Catalog of star observations for fgcm",
        name="fgcmStarObservations",
        storageClass="Catalog",
        dimensions=("instrument",),
    )
 
    fgcmStarIds = connectionTypes.Output(
        doc="Catalog of fgcm calibration star IDs",
        name="fgcmStarIds",
        storageClass="Catalog",
        dimensions=("instrument",),
    )
 
    fgcmStarIndices = connectionTypes.Output(
        doc="Catalog of fgcm calibration star indices",
        name="fgcmStarIndices",
        storageClass="Catalog",
        dimensions=("instrument",),
    )
 
    fgcmReferenceStars = connectionTypes.Output(
        doc="Catalog of fgcm-matched reference stars",
        name="fgcmReferenceStars",
        storageClass="Catalog",
        dimensions=("instrument",),
    )
 
    def __init__(self, *, config=None):
        super().__init__(config=config)
 
        if not config.doReferenceMatches:
            self.prerequisiteInputs.remove("refCat")
            self.prerequisiteInputs.remove("fgcmLookUpTable")
 
        if not config.doModelErrorsWithBackground:
            self.inputs.remove("background")
 
        if not config.doReferenceMatches:
            self.outputs.remove("fgcmReferenceStars")
 
 
class FgcmBuildStarsTableConfig(FgcmBuildStarsConfigBase, pipeBase.PipelineTaskConfig,
                                pipelineConnections=FgcmBuildStarsTableConnections):
    """Config for FgcmBuildStarsTableTask"""
 
    referenceCCD = pexConfig.Field(
        doc="Reference CCD for checking PSF and background",
        dtype=int,
        default=40,
    )
 
    def setDefaults(self):
        super().setDefaults()
 
        # The names here correspond to the post-transformed
        # sourceTable_visit catalogs, which differ from the raw src
        # catalogs.  Therefore, all field and flag names cannot
        # be derived from the base config class.
        self.instFluxFieldinstFluxFieldinstFluxField = 'ApFlux_12_0_instFlux'
        self.localBackgroundFluxFieldlocalBackgroundFluxFieldlocalBackgroundFluxField = 'LocalBackground_instFlux'
        self.apertureInnerInstFluxFieldapertureInnerInstFluxFieldapertureInnerInstFluxField = 'ApFlux_12_0_instFlux'
        self.apertureOuterInstFluxFieldapertureOuterInstFluxFieldapertureOuterInstFluxField = 'ApFlux_17_0_instFlux'
        self.psfCandidateNamepsfCandidateNamepsfCandidateName = 'Calib_psf_candidate'
 
        sourceSelector = self.sourceSelectorsourceSelector["science"]
 
        fluxFlagName = self.instFluxFieldinstFluxFieldinstFluxField[0: -len('instFlux')] + 'flag'
 
        sourceSelector.flags.bad = ['PixelFlags_edge',
                                    'PixelFlags_interpolatedCenter',
                                    'PixelFlags_saturatedCenter',
                                    'PixelFlags_crCenter',
                                    'PixelFlags_bad',
                                    'PixelFlags_interpolated',
                                    'PixelFlags_saturated',
                                    'Centroid_flag',
                                    fluxFlagName]
 
        if self.doSubtractLocalBackgrounddoSubtractLocalBackground:
            localBackgroundFlagName = self.localBackgroundFluxFieldlocalBackgroundFluxFieldlocalBackgroundFluxField[0: -len('instFlux')] + 'flag'
            sourceSelector.flags.bad.append(localBackgroundFlagName)
 
        sourceSelector.signalToNoise.fluxField = self.instFluxFieldinstFluxFieldinstFluxField
        sourceSelector.signalToNoise.errField = self.instFluxFieldinstFluxFieldinstFluxField + 'Err'
 
        sourceSelector.isolated.parentName = 'parentSourceId'
        sourceSelector.isolated.nChildName = 'Deblend_nChild'
 
        sourceSelector.unresolved.name = 'extendedness'
 
 
class FgcmBuildStarsTableTask(FgcmBuildStarsBaseTask):
    """
    Build stars for the FGCM global calibration, using sourceTable_visit catalogs.
    """
    ConfigClass = FgcmBuildStarsTableConfig
    RunnerClass = FgcmBuildStarsRunner
    _DefaultName = "fgcmBuildStarsTable"
 
    canMultiprocess = False
 
    def runQuantum(self, butlerQC, inputRefs, outputRefs):
        inputRefDict = butlerQC.get(inputRefs)
 
        sourceTableRefs = inputRefDict['sourceTable_visit']
 
        self.log.info("Running with %d sourceTable_visit dataRefs",
                      len(sourceTableRefs))
 
        sourceTableDataRefDict = {sourceTableRef.dataId['visit']: sourceTableRef for
                                  sourceTableRef in sourceTableRefs}
 
        if self.config.doReferenceMatches:
            # Get the LUT dataRef
            lutDataRef = inputRefDict['fgcmLookUpTable']
 
            # Prepare the refCat loader
            refConfig = self.config.fgcmLoadReferenceCatalog.refObjLoader
            refObjLoader = ReferenceObjectLoader(dataIds=[ref.datasetRef.dataId
                                                          for ref in inputRefs.refCat],
                                                 refCats=butlerQC.get(inputRefs.refCat),
                                                 config=refConfig,
                                                 log=self.log)
            self.makeSubtask('fgcmLoadReferenceCatalog', refObjLoader=refObjLoader)
        else:
            lutDataRef = None
 
        # Compute aperture radius if necessary.  This is useful to do now before
        # any heave lifting has happened (fail early).
        calibFluxApertureRadius = None
        if self.config.doSubtractLocalBackground:
            try:
                calibFluxApertureRadius = computeApertureRadiusFromDataRef(sourceTableRefs[0],
                                                                           self.config.instFluxField)
            except RuntimeError as e:
                raise RuntimeError("Could not determine aperture radius from %s. "
                                   "Cannot use doSubtractLocalBackground." %
                                   (self.config.instFluxField)) from e
 
        visitSummaryRefs = inputRefDict['visitSummary']
        visitSummaryDataRefDict = {visitSummaryRef.dataId['visit']: visitSummaryRef for
                                   visitSummaryRef in visitSummaryRefs}
 
        camera = inputRefDict['camera']
        groupedDataRefs = self._groupDataRefs_groupDataRefs(sourceTableDataRefDict,
                                              visitSummaryDataRefDict)
 
        if self.config.doModelErrorsWithBackground:
            bkgRefs = inputRefDict['background']
            bkgDataRefDict = {(bkgRef.dataId.byName()['visit'],
                               bkgRef.dataId.byName()['detector']): bkgRef for
                              bkgRef in bkgRefs}
        else:
            bkgDataRefDict = None
 
        # Gen3 does not currently allow "checkpoint" saving of datasets,
        # so we need to have this all in one go.
        visitCat = self.fgcmMakeVisitCatalogfgcmMakeVisitCatalog(camera, groupedDataRefs,
                                             bkgDataRefDict=bkgDataRefDict,
                                             visitCatDataRef=None,
                                             inVisitCat=None)
 
        rad = calibFluxApertureRadius
        sourceSchemaDataRef = inputRefDict['sourceSchema']
        fgcmStarObservationCat = self.fgcmMakeAllStarObservationsfgcmMakeAllStarObservationsfgcmMakeAllStarObservations(groupedDataRefs,
                                                                  visitCat,
                                                                  sourceSchemaDataRef,
                                                                  camera,
                                                                  calibFluxApertureRadius=rad,
                                                                  starObsDataRef=None,
                                                                  visitCatDataRef=None,
                                                                  inStarObsCat=None)
 
        butlerQC.put(visitCat, outputRefs.fgcmVisitCatalog)
        butlerQC.put(fgcmStarObservationCat, outputRefs.fgcmStarObservations)
 
        fgcmStarIdCat, fgcmStarIndicesCat, fgcmRefCat = self.fgcmMatchStarsfgcmMatchStars(visitCat,
                                                                            fgcmStarObservationCat,
                                                                            lutDataRef=lutDataRef)
 
        butlerQC.put(fgcmStarIdCat, outputRefs.fgcmStarIds)
        butlerQC.put(fgcmStarIndicesCat, outputRefs.fgcmStarIndices)
        if fgcmRefCat is not None:
            butlerQC.put(fgcmRefCat, outputRefs.fgcmReferenceStars)
 
    @classmethod
    def _makeArgumentParser(cls):
        """Create an argument parser"""
        parser = pipeBase.ArgumentParser(name=cls._DefaultName_DefaultName)
        parser.add_id_argument("--id", "sourceTable_visit", help="Data ID, e.g. --id visit=6789")
 
        return parser
 
    def _groupDataRefs(self, sourceTableDataRefDict, visitSummaryDataRefDict):
        """Group sourceTable and visitSummary dataRefs (gen3 only).
 
        Parameters
        ----------
        sourceTableDataRefDict : `dict` [`int`, `str`]
            Dict of source tables, keyed by visit.
        visitSummaryDataRefDict : `dict` [int, `str`]
            Dict of visit summary catalogs, keyed by visit.
 
        Returns
        -------
        groupedDataRefs : `dict` [`int`, `list`]
            Dictionary with sorted visit keys, and `list`s with
            `lsst.daf.butler.DeferredDataSetHandle`.  The first
            item in the list will be the visitSummary ref, and
            the second will be the source table ref.
        """
        groupedDataRefs = collections.defaultdict(list)
        visits = sorted(sourceTableDataRefDict.keys())
 
        for visit in visits:
            groupedDataRefs[visit] = [visitSummaryDataRefDict[visit],
                                      sourceTableDataRefDict[visit]]
 
        return groupedDataRefs
 
    def _findAndGroupDataRefsGen2(self, butler, camera, dataRefs):
        self.log.info("Grouping dataRefs by %s", (self.config.visitDataRefName))
 
        ccdIds = []
        for detector in camera:
            ccdIds.append(detector.getId())
        # Insert our preferred referenceCCD first:
        # It is fine that this is listed twice, because we only need
        # the first calexp that is found.
        ccdIds.insert(0, self.config.referenceCCD)
 
        # The visitTable building code expects a dictionary of groupedDataRefs
        # keyed by visit, the first element as the "primary" calexp dataRef.
        # We then append the sourceTable_visit dataRef at the end for the
        # code which does the data reading (fgcmMakeAllStarObservations).
 
        groupedDataRefs = collections.defaultdict(list)
        for dataRef in dataRefs:
            visit = dataRef.dataId[self.config.visitDataRefName]
 
            # Find an existing calexp (we need for psf and metadata)
            # and make the relevant dataRef
            for ccdId in ccdIds:
                try:
                    calexpRef = butler.dataRef('calexp', dataId={self.config.visitDataRefName: visit,
                                                                 self.config.ccdDataRefName: ccdId})
                except RuntimeError:
                    # Not found
                    continue
 
                # Make sure the dataset exists
                if not calexpRef.datasetExists():
                    continue
 
                # It was found.  Add and quit out, since we only
                # need one calexp per visit.
                groupedDataRefs[visit].append(calexpRef)
                break
 
            # And append this dataRef
            groupedDataRefs[visit].append(dataRef)
 
        # This should be sorted by visit (the key)
        return dict(sorted(groupedDataRefs.items()))
 
    def fgcmMakeAllStarObservations(self, groupedDataRefs, visitCat,
                                    sourceSchemaDataRef,
                                    camera,
                                    calibFluxApertureRadius=None,
                                    visitCatDataRef=None,
                                    starObsDataRef=None,
                                    inStarObsCat=None):
        startTime = time.time()
 
        # If both dataRefs are None, then we assume the caller does not
        # want to store checkpoint files.  If both are set, we will
        # do checkpoint files.  And if only one is set, this is potentially
        # unintentional and we will warn.
        if (visitCatDataRef is not None and starObsDataRef is None
           or visitCatDataRef is None and starObsDataRef is not None):
            self.log.warn("Only one of visitCatDataRef and starObsDataRef are set, so "
                          "no checkpoint files will be persisted.")
 
        if self.config.doSubtractLocalBackground and calibFluxApertureRadius is None:
            raise RuntimeError("Must set calibFluxApertureRadius if doSubtractLocalBackground is True.")
 
        # To get the correct output schema, we use similar code as fgcmBuildStarsTask
        # We are not actually using this mapper, except to grab the outputSchema
        sourceSchema = sourceSchemaDataRef.get().schema
        sourceMapper = self._makeSourceMapper_makeSourceMapper(sourceSchema)
        outputSchema = sourceMapper.getOutputSchema()
 
        # Construct mapping from ccd number to index
        ccdMapping = {}
        for ccdIndex, detector in enumerate(camera):
            ccdMapping[detector.getId()] = ccdIndex
 
        approxPixelAreaFields = computeApproxPixelAreaFields(camera)
 
        if inStarObsCat is not None:
            fullCatalog = inStarObsCat
            comp1 = fullCatalog.schema.compare(outputSchema, outputSchema.EQUAL_KEYS)
            comp2 = fullCatalog.schema.compare(outputSchema, outputSchema.EQUAL_NAMES)
            if not comp1 or not comp2:
                raise RuntimeError("Existing fgcmStarObservations file found with mismatched schema.")
        else:
            fullCatalog = afwTable.BaseCatalog(outputSchema)
 
        visitKey = outputSchema['visit'].asKey()
        ccdKey = outputSchema['ccd'].asKey()
        instMagKey = outputSchema['instMag'].asKey()
        instMagErrKey = outputSchema['instMagErr'].asKey()
 
        # Prepare local background if desired
        if self.config.doSubtractLocalBackground:
            localBackgroundArea = np.pi*calibFluxApertureRadius**2.
 
        # Determine which columns we need from the sourceTable_visit catalogs
        columns = self._get_sourceTable_visit_columns_get_sourceTable_visit_columns()
 
        k = 2.5/np.log(10.)
 
        for counter, visit in enumerate(visitCat):
            # Check if these sources have already been read and stored in the checkpoint file
            if visit['sources_read']:
                continue
 
            expTime = visit['exptime']
 
            dataRef = groupedDataRefs[visit['visit']][-1]
 
            if isinstance(dataRef, dafPersist.ButlerDataRef):
                srcTable = dataRef.get()
                df = srcTable.toDataFrame(columns)
            else:
                df = dataRef.get(parameters={'columns': columns})
 
            goodSrc = self.sourceSelector.selectSources(df)
 
            # Need to add a selection based on the local background correction
            # if necessary
            if self.config.doSubtractLocalBackground:
                localBackground = localBackgroundArea*df[self.config.localBackgroundFluxField].values
                use, = np.where((goodSrc.selected)
                                & ((df[self.config.instFluxField].values - localBackground) > 0.0))
            else:
                use, = np.where(goodSrc.selected)
 
            tempCat = afwTable.BaseCatalog(fullCatalog.schema)
            tempCat.resize(use.size)
 
            tempCat['ra'][:] = np.deg2rad(df['ra'].values[use])
            tempCat['dec'][:] = np.deg2rad(df['decl'].values[use])
            tempCat['x'][:] = df['x'].values[use]
            tempCat['y'][:] = df['y'].values[use]
            # These "visit" and "ccd" names in the parquet tables are
            # hard-coded.
            tempCat[visitKey][:] = df['visit'].values[use]
            tempCat[ccdKey][:] = df['ccd'].values[use]
            tempCat['psf_candidate'] = df['Calib_psf_candidate'].values[use]
 
            if self.config.doSubtractLocalBackground:
                # At the moment we only adjust the flux and not the flux
                # error by the background because the error on
                # base_LocalBackground_instFlux is the rms error in the
                # background annulus, not the error on the mean in the
                # background estimate (which is much smaller, by sqrt(n)
                # pixels used to estimate the background, which we do not
                # have access to in this task).  In the default settings,
                # the annulus is sufficiently large such that these
                # additional errors are are negligibly small (much less
                # than a mmag in quadrature).
 
                # This is the difference between the mag with local background correction
                # and the mag without local background correction.
                tempCat['deltaMagBkg'] = (-2.5*np.log10(df[self.config.instFluxField].values[use]
                                                        - localBackground[use]) -
                                          -2.5*np.log10(df[self.config.instFluxField].values[use]))
            else:
                tempCat['deltaMagBkg'][:] = 0.0
 
            # Need to loop over ccds here
            for detector in camera:
                ccdId = detector.getId()
                # used index for all observations with a given ccd
                use2 = (tempCat[ccdKey] == ccdId)
                tempCat['jacobian'][use2] = approxPixelAreaFields[ccdId].evaluate(tempCat['x'][use2],
                                                                                  tempCat['y'][use2])
                scaledInstFlux = (df[self.config.instFluxField].values[use[use2]]
                                  * visit['scaling'][ccdMapping[ccdId]])
                tempCat[instMagKey][use2] = (-2.5*np.log10(scaledInstFlux) + 2.5*np.log10(expTime))
 
            # Compute instMagErr from instFluxErr/instFlux, any scaling
            # will cancel out.
            tempCat[instMagErrKey][:] = k*(df[self.config.instFluxField + 'Err'].values[use]
                                           / df[self.config.instFluxField].values[use])
 
            # Apply the jacobian if configured
            if self.config.doApplyWcsJacobian:
                tempCat[instMagKey][:] -= 2.5*np.log10(tempCat['jacobian'][:])
 
            fullCatalog.extend(tempCat)
 
            # Now do the aperture information
            with np.warnings.catch_warnings():
                # Ignore warnings, we will filter infinites and nans below
                np.warnings.simplefilter("ignore")
 
                instMagIn = -2.5*np.log10(df[self.config.apertureInnerInstFluxField].values[use])
                instMagErrIn = k*(df[self.config.apertureInnerInstFluxField + 'Err'].values[use]
                                  / df[self.config.apertureInnerInstFluxField].values[use])
                instMagOut = -2.5*np.log10(df[self.config.apertureOuterInstFluxField].values[use])
                instMagErrOut = k*(df[self.config.apertureOuterInstFluxField + 'Err'].values[use]
                                   / df[self.config.apertureOuterInstFluxField].values[use])
 
            ok = (np.isfinite(instMagIn) & np.isfinite(instMagErrIn)
                  & np.isfinite(instMagOut) & np.isfinite(instMagErrOut))
 
            visit['deltaAper'] = np.median(instMagIn[ok] - instMagOut[ok])
            visit['sources_read'] = True
 
            self.log.info("  Found %d good stars in visit %d (deltaAper = %0.3f)",
                          use.size, visit['visit'], visit['deltaAper'])
 
            if ((counter % self.config.nVisitsPerCheckpoint) == 0
               and starObsDataRef is not None and visitCatDataRef is not None):
                # We need to persist both the stars and the visit catalog which gets
                # additional metadata from each visit.
                starObsDataRef.put(fullCatalog)
                visitCatDataRef.put(visitCat)
 
        self.log.info("Found all good star observations in %.2f s" %
                      (time.time() - startTime))
 
        return fullCatalog
 
    def _get_sourceTable_visit_columns(self):
        """
        Get the sourceTable_visit columns from the config.
 
        Returns
        -------
        columns : `list`
           List of columns to read from sourceTable_visit
        """
        # These "visit" and "ccd" names in the parquet tables are hard-coded.
        columns = ['visit', 'ccd',
                   'ra', 'decl', 'x', 'y', self.config.psfCandidateName,
                   self.config.instFluxField, self.config.instFluxField + 'Err',
                   self.config.apertureInnerInstFluxField, self.config.apertureInnerInstFluxField + 'Err',
                   self.config.apertureOuterInstFluxField, self.config.apertureOuterInstFluxField + 'Err']
        if self.sourceSelector.config.doFlags:
            columns.extend(self.sourceSelector.config.flags.bad)
        if self.sourceSelector.config.doUnresolved:
            columns.append(self.sourceSelector.config.unresolved.name)
        if self.sourceSelector.config.doIsolated:
            columns.append(self.sourceSelector.config.isolated.parentName)
            columns.append(self.sourceSelector.config.isolated.nChildName)
        if self.config.doSubtractLocalBackground:
            columns.append(self.config.localBackgroundFluxField)
 
        return columns