Coverage for python/lsst/ip/isr/isrTask.py : 16%

Hot-keys on this page
r m x p toggle line displays
j k next/prev highlighted chunk
0 (zero) top of page
1 (one) first highlighted chunk
1# This file is part of ip_isr.
2#
3# Developed for the LSST Data Management System.
4# This product includes software developed by the LSST Project
5# (https://www.lsst.org).
6# See the COPYRIGHT file at the top-level directory of this distribution
7# for details of code ownership.
8#
9# This program is free software: you can redistribute it and/or modify
10# it under the terms of the GNU General Public License as published by
11# the Free Software Foundation, either version 3 of the License, or
12# (at your option) any later version.
13#
14# This program is distributed in the hope that it will be useful,
15# but WITHOUT ANY WARRANTY; without even the implied warranty of
16# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
17# GNU General Public License for more details.
18#
19# You should have received a copy of the GNU General Public License
20# along with this program. If not, see <https://www.gnu.org/licenses/>.
22import math
23import numpy
25import lsst.geom
26import lsst.afw.image as afwImage
27import lsst.afw.math as afwMath
28import lsst.pex.config as pexConfig
29import lsst.pipe.base as pipeBase
30import lsst.pipe.base.connectionTypes as cT
32from contextlib import contextmanager
33from lsstDebug import getDebugFrame
35from lsst.afw.cameraGeom import (PIXELS, FOCAL_PLANE, NullLinearityType,
36 ReadoutCorner)
37from lsst.afw.display import getDisplay
38from lsst.afw.geom import Polygon
39from lsst.daf.persistence import ButlerDataRef
40from lsst.daf.persistence.butler import NoResults
41from lsst.meas.algorithms.detection import SourceDetectionTask
43from . import isrFunctions
44from . import isrQa
45from . import linearize
46from .defects import Defects
48from .assembleCcdTask import AssembleCcdTask
49from .crosstalk import CrosstalkTask, CrosstalkCalib
50from .fringe import FringeTask
51from .isr import maskNans
52from .masking import MaskingTask
53from .overscan import OverscanCorrectionTask
54from .straylight import StrayLightTask
55from .vignette import VignetteTask
56from .ampOffset import AmpOffsetTask
57from lsst.daf.butler import DimensionGraph
60__all__ = ["IsrTask", "IsrTaskConfig", "RunIsrTask", "RunIsrConfig"]
63def crosstalkSourceLookup(datasetType, registry, quantumDataId, collections):
64 """Lookup function to identify crosstalkSource entries.
66 This should return an empty list under most circumstances. Only
67 when inter-chip crosstalk has been identified should this be
68 populated.
70 This will be unused until DM-25348 resolves the quantum graph
71 generation issue.
73 Parameters
74 ----------
75 datasetType : `str`
76 Dataset to lookup.
77 registry : `lsst.daf.butler.Registry`
78 Butler registry to query.
79 quantumDataId : `lsst.daf.butler.ExpandedDataCoordinate`
80 Data id to transform to identify crosstalkSources. The
81 ``detector`` entry will be stripped.
82 collections : `lsst.daf.butler.CollectionSearch`
83 Collections to search through.
85 Returns
86 -------
87 results : `list` [`lsst.daf.butler.DatasetRef`]
88 List of datasets that match the query that will be used as
89 crosstalkSources.
90 """
91 newDataId = quantumDataId.subset(DimensionGraph(registry.dimensions, names=["instrument", "exposure"]))
92 results = list(registry.queryDatasets(datasetType,
93 collections=collections,
94 dataId=newDataId,
95 findFirst=True,
96 ).expanded())
97 return results
100class IsrTaskConnections(pipeBase.PipelineTaskConnections,
101 dimensions={"instrument", "exposure", "detector"},
102 defaultTemplates={}):
103 ccdExposure = cT.Input(
104 name="raw",
105 doc="Input exposure to process.",
106 storageClass="Exposure",
107 dimensions=["instrument", "exposure", "detector"],
108 )
109 camera = cT.PrerequisiteInput(
110 name="camera",
111 storageClass="Camera",
112 doc="Input camera to construct complete exposures.",
113 dimensions=["instrument"],
114 isCalibration=True,
115 )
117 crosstalk = cT.PrerequisiteInput(
118 name="crosstalk",
119 doc="Input crosstalk object",
120 storageClass="CrosstalkCalib",
121 dimensions=["instrument", "detector"],
122 isCalibration=True,
123 minimum=0, # can fall back to cameraGeom
124 )
125 # TODO: DM-25348. This does not work yet to correctly load
126 # possible crosstalk sources.
127 crosstalkSources = cT.PrerequisiteInput(
128 name="isrOverscanCorrected",
129 doc="Overscan corrected input images.",
130 storageClass="Exposure",
131 dimensions=["instrument", "exposure", "detector"],
132 deferLoad=True,
133 multiple=True,
134 lookupFunction=crosstalkSourceLookup,
135 minimum=0, # not needed for all instruments, no config to control this
136 )
137 bias = cT.PrerequisiteInput(
138 name="bias",
139 doc="Input bias calibration.",
140 storageClass="ExposureF",
141 dimensions=["instrument", "detector"],
142 isCalibration=True,
143 )
144 dark = cT.PrerequisiteInput(
145 name='dark',
146 doc="Input dark calibration.",
147 storageClass="ExposureF",
148 dimensions=["instrument", "detector"],
149 isCalibration=True,
150 )
151 flat = cT.PrerequisiteInput(
152 name="flat",
153 doc="Input flat calibration.",
154 storageClass="ExposureF",
155 dimensions=["instrument", "physical_filter", "detector"],
156 isCalibration=True,
157 )
158 ptc = cT.PrerequisiteInput(
159 name="ptc",
160 doc="Input Photon Transfer Curve dataset",
161 storageClass="PhotonTransferCurveDataset",
162 dimensions=["instrument", "detector"],
163 isCalibration=True,
164 )
165 fringes = cT.PrerequisiteInput(
166 name="fringe",
167 doc="Input fringe calibration.",
168 storageClass="ExposureF",
169 dimensions=["instrument", "physical_filter", "detector"],
170 isCalibration=True,
171 minimum=0, # only needed for some bands, even when enabled
172 )
173 strayLightData = cT.PrerequisiteInput(
174 name='yBackground',
175 doc="Input stray light calibration.",
176 storageClass="StrayLightData",
177 dimensions=["instrument", "physical_filter", "detector"],
178 deferLoad=True,
179 isCalibration=True,
180 minimum=0, # only needed for some bands, even when enabled
181 )
182 bfKernel = cT.PrerequisiteInput(
183 name='bfKernel',
184 doc="Input brighter-fatter kernel.",
185 storageClass="NumpyArray",
186 dimensions=["instrument"],
187 isCalibration=True,
188 minimum=0, # can use either bfKernel or newBFKernel
189 )
190 newBFKernel = cT.PrerequisiteInput(
191 name='brighterFatterKernel',
192 doc="Newer complete kernel + gain solutions.",
193 storageClass="BrighterFatterKernel",
194 dimensions=["instrument", "detector"],
195 isCalibration=True,
196 minimum=0, # can use either bfKernel or newBFKernel
197 )
198 defects = cT.PrerequisiteInput(
199 name='defects',
200 doc="Input defect tables.",
201 storageClass="Defects",
202 dimensions=["instrument", "detector"],
203 isCalibration=True,
204 )
205 linearizer = cT.PrerequisiteInput(
206 name='linearizer',
207 storageClass="Linearizer",
208 doc="Linearity correction calibration.",
209 dimensions=["instrument", "detector"],
210 isCalibration=True,
211 minimum=0, # can fall back to cameraGeom
212 )
213 opticsTransmission = cT.PrerequisiteInput(
214 name="transmission_optics",
215 storageClass="TransmissionCurve",
216 doc="Transmission curve due to the optics.",
217 dimensions=["instrument"],
218 isCalibration=True,
219 )
220 filterTransmission = cT.PrerequisiteInput(
221 name="transmission_filter",
222 storageClass="TransmissionCurve",
223 doc="Transmission curve due to the filter.",
224 dimensions=["instrument", "physical_filter"],
225 isCalibration=True,
226 )
227 sensorTransmission = cT.PrerequisiteInput(
228 name="transmission_sensor",
229 storageClass="TransmissionCurve",
230 doc="Transmission curve due to the sensor.",
231 dimensions=["instrument", "detector"],
232 isCalibration=True,
233 )
234 atmosphereTransmission = cT.PrerequisiteInput(
235 name="transmission_atmosphere",
236 storageClass="TransmissionCurve",
237 doc="Transmission curve due to the atmosphere.",
238 dimensions=["instrument"],
239 isCalibration=True,
240 )
241 illumMaskedImage = cT.PrerequisiteInput(
242 name="illum",
243 doc="Input illumination correction.",
244 storageClass="MaskedImageF",
245 dimensions=["instrument", "physical_filter", "detector"],
246 isCalibration=True,
247 )
249 outputExposure = cT.Output(
250 name='postISRCCD',
251 doc="Output ISR processed exposure.",
252 storageClass="Exposure",
253 dimensions=["instrument", "exposure", "detector"],
254 )
255 preInterpExposure = cT.Output(
256 name='preInterpISRCCD',
257 doc="Output ISR processed exposure, with pixels left uninterpolated.",
258 storageClass="ExposureF",
259 dimensions=["instrument", "exposure", "detector"],
260 )
261 outputOssThumbnail = cT.Output(
262 name="OssThumb",
263 doc="Output Overscan-subtracted thumbnail image.",
264 storageClass="Thumbnail",
265 dimensions=["instrument", "exposure", "detector"],
266 )
267 outputFlattenedThumbnail = cT.Output(
268 name="FlattenedThumb",
269 doc="Output flat-corrected thumbnail image.",
270 storageClass="Thumbnail",
271 dimensions=["instrument", "exposure", "detector"],
272 )
274 def __init__(self, *, config=None):
275 super().__init__(config=config)
277 if config.doBias is not True:
278 self.prerequisiteInputs.discard("bias")
279 if config.doLinearize is not True:
280 self.prerequisiteInputs.discard("linearizer")
281 if config.doCrosstalk is not True:
282 self.inputs.discard("crosstalkSources")
283 self.prerequisiteInputs.discard("crosstalk")
284 if config.doBrighterFatter is not True:
285 self.prerequisiteInputs.discard("bfKernel")
286 self.prerequisiteInputs.discard("newBFKernel")
287 if config.doDefect is not True:
288 self.prerequisiteInputs.discard("defects")
289 if config.doDark is not True:
290 self.prerequisiteInputs.discard("dark")
291 if config.doFlat is not True:
292 self.prerequisiteInputs.discard("flat")
293 if config.doFringe is not True:
294 self.prerequisiteInputs.discard("fringe")
295 if config.doStrayLight is not True:
296 self.prerequisiteInputs.discard("strayLightData")
297 if config.usePtcGains is not True and config.usePtcReadNoise is not True:
298 self.prerequisiteInputs.discard("ptc")
299 if config.doAttachTransmissionCurve is not True:
300 self.prerequisiteInputs.discard("opticsTransmission")
301 self.prerequisiteInputs.discard("filterTransmission")
302 self.prerequisiteInputs.discard("sensorTransmission")
303 self.prerequisiteInputs.discard("atmosphereTransmission")
304 if config.doUseOpticsTransmission is not True:
305 self.prerequisiteInputs.discard("opticsTransmission")
306 if config.doUseFilterTransmission is not True:
307 self.prerequisiteInputs.discard("filterTransmission")
308 if config.doUseSensorTransmission is not True:
309 self.prerequisiteInputs.discard("sensorTransmission")
310 if config.doUseAtmosphereTransmission is not True:
311 self.prerequisiteInputs.discard("atmosphereTransmission")
312 if config.doIlluminationCorrection is not True:
313 self.prerequisiteInputs.discard("illumMaskedImage")
315 if config.doWrite is not True:
316 self.outputs.discard("outputExposure")
317 self.outputs.discard("preInterpExposure")
318 self.outputs.discard("outputFlattenedThumbnail")
319 self.outputs.discard("outputOssThumbnail")
320 if config.doSaveInterpPixels is not True:
321 self.outputs.discard("preInterpExposure")
322 if config.qa.doThumbnailOss is not True:
323 self.outputs.discard("outputOssThumbnail")
324 if config.qa.doThumbnailFlattened is not True:
325 self.outputs.discard("outputFlattenedThumbnail")
328class IsrTaskConfig(pipeBase.PipelineTaskConfig,
329 pipelineConnections=IsrTaskConnections):
330 """Configuration parameters for IsrTask.
332 Items are grouped in the order in which they are executed by the task.
333 """
334 datasetType = pexConfig.Field(
335 dtype=str,
336 doc="Dataset type for input data; users will typically leave this alone, "
337 "but camera-specific ISR tasks will override it",
338 default="raw",
339 )
341 fallbackFilterName = pexConfig.Field(
342 dtype=str,
343 doc="Fallback default filter name for calibrations.",
344 optional=True
345 )
346 useFallbackDate = pexConfig.Field(
347 dtype=bool,
348 doc="Pass observation date when using fallback filter.",
349 default=False,
350 )
351 expectWcs = pexConfig.Field(
352 dtype=bool,
353 default=True,
354 doc="Expect input science images to have a WCS (set False for e.g. spectrographs)."
355 )
356 fwhm = pexConfig.Field(
357 dtype=float,
358 doc="FWHM of PSF in arcseconds.",
359 default=1.0,
360 )
361 qa = pexConfig.ConfigField(
362 dtype=isrQa.IsrQaConfig,
363 doc="QA related configuration options.",
364 )
366 # Image conversion configuration
367 doConvertIntToFloat = pexConfig.Field(
368 dtype=bool,
369 doc="Convert integer raw images to floating point values?",
370 default=True,
371 )
373 # Saturated pixel handling.
374 doSaturation = pexConfig.Field(
375 dtype=bool,
376 doc="Mask saturated pixels? NB: this is totally independent of the"
377 " interpolation option - this is ONLY setting the bits in the mask."
378 " To have them interpolated make sure doSaturationInterpolation=True",
379 default=True,
380 )
381 saturatedMaskName = pexConfig.Field(
382 dtype=str,
383 doc="Name of mask plane to use in saturation detection and interpolation",
384 default="SAT",
385 )
386 saturation = pexConfig.Field(
387 dtype=float,
388 doc="The saturation level to use if no Detector is present in the Exposure (ignored if NaN)",
389 default=float("NaN"),
390 )
391 growSaturationFootprintSize = pexConfig.Field(
392 dtype=int,
393 doc="Number of pixels by which to grow the saturation footprints",
394 default=1,
395 )
397 # Suspect pixel handling.
398 doSuspect = pexConfig.Field(
399 dtype=bool,
400 doc="Mask suspect pixels?",
401 default=False,
402 )
403 suspectMaskName = pexConfig.Field(
404 dtype=str,
405 doc="Name of mask plane to use for suspect pixels",
406 default="SUSPECT",
407 )
408 numEdgeSuspect = pexConfig.Field(
409 dtype=int,
410 doc="Number of edge pixels to be flagged as untrustworthy.",
411 default=0,
412 )
413 edgeMaskLevel = pexConfig.ChoiceField(
414 dtype=str,
415 doc="Mask edge pixels in which coordinate frame: DETECTOR or AMP?",
416 default="DETECTOR",
417 allowed={
418 'DETECTOR': 'Mask only the edges of the full detector.',
419 'AMP': 'Mask edges of each amplifier.',
420 },
421 )
423 # Initial masking options.
424 doSetBadRegions = pexConfig.Field(
425 dtype=bool,
426 doc="Should we set the level of all BAD patches of the chip to the chip's average value?",
427 default=True,
428 )
429 badStatistic = pexConfig.ChoiceField(
430 dtype=str,
431 doc="How to estimate the average value for BAD regions.",
432 default='MEANCLIP',
433 allowed={
434 "MEANCLIP": "Correct using the (clipped) mean of good data",
435 "MEDIAN": "Correct using the median of the good data",
436 },
437 )
439 # Overscan subtraction configuration.
440 doOverscan = pexConfig.Field(
441 dtype=bool,
442 doc="Do overscan subtraction?",
443 default=True,
444 )
445 overscan = pexConfig.ConfigurableField(
446 target=OverscanCorrectionTask,
447 doc="Overscan subtraction task for image segments.",
448 )
449 overscanFitType = pexConfig.ChoiceField(
450 dtype=str,
451 doc="The method for fitting the overscan bias level.",
452 default='MEDIAN',
453 allowed={
454 "POLY": "Fit ordinary polynomial to the longest axis of the overscan region",
455 "CHEB": "Fit Chebyshev polynomial to the longest axis of the overscan region",
456 "LEG": "Fit Legendre polynomial to the longest axis of the overscan region",
457 "NATURAL_SPLINE": "Fit natural spline to the longest axis of the overscan region",
458 "CUBIC_SPLINE": "Fit cubic spline to the longest axis of the overscan region",
459 "AKIMA_SPLINE": "Fit Akima spline to the longest axis of the overscan region",
460 "MEAN": "Correct using the mean of the overscan region",
461 "MEANCLIP": "Correct using a clipped mean of the overscan region",
462 "MEDIAN": "Correct using the median of the overscan region",
463 "MEDIAN_PER_ROW": "Correct using the median per row of the overscan region",
464 },
465 deprecated=("Please configure overscan via the OverscanCorrectionConfig interface."
466 " This option will no longer be used, and will be removed after v20.")
467 )
468 overscanOrder = pexConfig.Field(
469 dtype=int,
470 doc=("Order of polynomial or to fit if overscan fit type is a polynomial, "
471 "or number of spline knots if overscan fit type is a spline."),
472 default=1,
473 deprecated=("Please configure overscan via the OverscanCorrectionConfig interface."
474 " This option will no longer be used, and will be removed after v20.")
475 )
476 overscanNumSigmaClip = pexConfig.Field(
477 dtype=float,
478 doc="Rejection threshold (sigma) for collapsing overscan before fit",
479 default=3.0,
480 deprecated=("Please configure overscan via the OverscanCorrectionConfig interface."
481 " This option will no longer be used, and will be removed after v20.")
482 )
483 overscanIsInt = pexConfig.Field(
484 dtype=bool,
485 doc="Treat overscan as an integer image for purposes of overscan.FitType=MEDIAN"
486 " and overscan.FitType=MEDIAN_PER_ROW.",
487 default=True,
488 deprecated=("Please configure overscan via the OverscanCorrectionConfig interface."
489 " This option will no longer be used, and will be removed after v20.")
490 )
491 # These options do not get deprecated, as they define how we slice up the
492 # image data.
493 overscanNumLeadingColumnsToSkip = pexConfig.Field(
494 dtype=int,
495 doc="Number of columns to skip in overscan, i.e. those closest to amplifier",
496 default=0,
497 )
498 overscanNumTrailingColumnsToSkip = pexConfig.Field(
499 dtype=int,
500 doc="Number of columns to skip in overscan, i.e. those farthest from amplifier",
501 default=0,
502 )
503 overscanMaxDev = pexConfig.Field( 503 ↛ exitline 503 didn't jump to the function exit
504 dtype=float,
505 doc="Maximum deviation from the median for overscan",
506 default=1000.0, check=lambda x: x > 0
507 )
508 overscanBiasJump = pexConfig.Field(
509 dtype=bool,
510 doc="Fit the overscan in a piecewise-fashion to correct for bias jumps?",
511 default=False,
512 )
513 overscanBiasJumpKeyword = pexConfig.Field(
514 dtype=str,
515 doc="Header keyword containing information about devices.",
516 default="NO_SUCH_KEY",
517 )
518 overscanBiasJumpDevices = pexConfig.ListField(
519 dtype=str,
520 doc="List of devices that need piecewise overscan correction.",
521 default=(),
522 )
523 overscanBiasJumpLocation = pexConfig.Field(
524 dtype=int,
525 doc="Location of bias jump along y-axis.",
526 default=0,
527 )
529 # Amplifier to CCD assembly configuration
530 doAssembleCcd = pexConfig.Field(
531 dtype=bool,
532 default=True,
533 doc="Assemble amp-level exposures into a ccd-level exposure?"
534 )
535 assembleCcd = pexConfig.ConfigurableField(
536 target=AssembleCcdTask,
537 doc="CCD assembly task",
538 )
540 # General calibration configuration.
541 doAssembleIsrExposures = pexConfig.Field(
542 dtype=bool,
543 default=False,
544 doc="Assemble amp-level calibration exposures into ccd-level exposure?"
545 )
546 doTrimToMatchCalib = pexConfig.Field(
547 dtype=bool,
548 default=False,
549 doc="Trim raw data to match calibration bounding boxes?"
550 )
552 # Bias subtraction.
553 doBias = pexConfig.Field(
554 dtype=bool,
555 doc="Apply bias frame correction?",
556 default=True,
557 )
558 biasDataProductName = pexConfig.Field(
559 dtype=str,
560 doc="Name of the bias data product",
561 default="bias",
562 )
563 doBiasBeforeOverscan = pexConfig.Field(
564 dtype=bool,
565 doc="Reverse order of overscan and bias correction.",
566 default=False
567 )
569 # Variance construction
570 doVariance = pexConfig.Field(
571 dtype=bool,
572 doc="Calculate variance?",
573 default=True
574 )
575 gain = pexConfig.Field(
576 dtype=float,
577 doc="The gain to use if no Detector is present in the Exposure (ignored if NaN)",
578 default=float("NaN"),
579 )
580 readNoise = pexConfig.Field(
581 dtype=float,
582 doc="The read noise to use if no Detector is present in the Exposure",
583 default=0.0,
584 )
585 doEmpiricalReadNoise = pexConfig.Field(
586 dtype=bool,
587 default=False,
588 doc="Calculate empirical read noise instead of value from AmpInfo data?"
589 )
590 usePtcReadNoise = pexConfig.Field(
591 dtype=bool,
592 default=False,
593 doc="Use readnoise values from the Photon Transfer Curve?"
594 )
595 maskNegativeVariance = pexConfig.Field(
596 dtype=bool,
597 default=True,
598 doc="Mask pixels that claim a negative variance? This likely indicates a failure "
599 "in the measurement of the overscan at an edge due to the data falling off faster "
600 "than the overscan model can account for it."
601 )
602 negativeVarianceMaskName = pexConfig.Field(
603 dtype=str,
604 default="BAD",
605 doc="Mask plane to use to mark pixels with negative variance, if `maskNegativeVariance` is True.",
606 )
607 # Linearization.
608 doLinearize = pexConfig.Field(
609 dtype=bool,
610 doc="Correct for nonlinearity of the detector's response?",
611 default=True,
612 )
614 # Crosstalk.
615 doCrosstalk = pexConfig.Field(
616 dtype=bool,
617 doc="Apply intra-CCD crosstalk correction?",
618 default=False,
619 )
620 doCrosstalkBeforeAssemble = pexConfig.Field(
621 dtype=bool,
622 doc="Apply crosstalk correction before CCD assembly, and before trimming?",
623 default=False,
624 )
625 crosstalk = pexConfig.ConfigurableField(
626 target=CrosstalkTask,
627 doc="Intra-CCD crosstalk correction",
628 )
630 # Masking options.
631 doDefect = pexConfig.Field(
632 dtype=bool,
633 doc="Apply correction for CCD defects, e.g. hot pixels?",
634 default=True,
635 )
636 doNanMasking = pexConfig.Field(
637 dtype=bool,
638 doc="Mask non-finite (NAN, inf) pixels?",
639 default=True,
640 )
641 doWidenSaturationTrails = pexConfig.Field(
642 dtype=bool,
643 doc="Widen bleed trails based on their width?",
644 default=True
645 )
647 # Brighter-Fatter correction.
648 doBrighterFatter = pexConfig.Field(
649 dtype=bool,
650 default=False,
651 doc="Apply the brighter-fatter correction?"
652 )
653 brighterFatterLevel = pexConfig.ChoiceField(
654 dtype=str,
655 default="DETECTOR",
656 doc="The level at which to correct for brighter-fatter.",
657 allowed={
658 "AMP": "Every amplifier treated separately.",
659 "DETECTOR": "One kernel per detector",
660 }
661 )
662 brighterFatterMaxIter = pexConfig.Field(
663 dtype=int,
664 default=10,
665 doc="Maximum number of iterations for the brighter-fatter correction"
666 )
667 brighterFatterThreshold = pexConfig.Field(
668 dtype=float,
669 default=1000,
670 doc="Threshold used to stop iterating the brighter-fatter correction. It is the "
671 "absolute value of the difference between the current corrected image and the one "
672 "from the previous iteration summed over all the pixels."
673 )
674 brighterFatterApplyGain = pexConfig.Field(
675 dtype=bool,
676 default=True,
677 doc="Should the gain be applied when applying the brighter-fatter correction?"
678 )
679 brighterFatterMaskListToInterpolate = pexConfig.ListField(
680 dtype=str,
681 doc="List of mask planes that should be interpolated over when applying the brighter-fatter "
682 "correction.",
683 default=["SAT", "BAD", "NO_DATA", "UNMASKEDNAN"],
684 )
685 brighterFatterMaskGrowSize = pexConfig.Field(
686 dtype=int,
687 default=0,
688 doc="Number of pixels to grow the masks listed in config.brighterFatterMaskListToInterpolate "
689 "when brighter-fatter correction is applied."
690 )
692 # Dark subtraction.
693 doDark = pexConfig.Field(
694 dtype=bool,
695 doc="Apply dark frame correction?",
696 default=True,
697 )
698 darkDataProductName = pexConfig.Field(
699 dtype=str,
700 doc="Name of the dark data product",
701 default="dark",
702 )
704 # Camera-specific stray light removal.
705 doStrayLight = pexConfig.Field(
706 dtype=bool,
707 doc="Subtract stray light in the y-band (due to encoder LEDs)?",
708 default=False,
709 )
710 strayLight = pexConfig.ConfigurableField(
711 target=StrayLightTask,
712 doc="y-band stray light correction"
713 )
715 # Flat correction.
716 doFlat = pexConfig.Field(
717 dtype=bool,
718 doc="Apply flat field correction?",
719 default=True,
720 )
721 flatDataProductName = pexConfig.Field(
722 dtype=str,
723 doc="Name of the flat data product",
724 default="flat",
725 )
726 flatScalingType = pexConfig.ChoiceField(
727 dtype=str,
728 doc="The method for scaling the flat on the fly.",
729 default='USER',
730 allowed={
731 "USER": "Scale by flatUserScale",
732 "MEAN": "Scale by the inverse of the mean",
733 "MEDIAN": "Scale by the inverse of the median",
734 },
735 )
736 flatUserScale = pexConfig.Field(
737 dtype=float,
738 doc="If flatScalingType is 'USER' then scale flat by this amount; ignored otherwise",
739 default=1.0,
740 )
741 doTweakFlat = pexConfig.Field(
742 dtype=bool,
743 doc="Tweak flats to match observed amplifier ratios?",
744 default=False
745 )
747 # Amplifier normalization based on gains instead of using flats
748 # configuration.
749 doApplyGains = pexConfig.Field(
750 dtype=bool,
751 doc="Correct the amplifiers for their gains instead of applying flat correction",
752 default=False,
753 )
754 usePtcGains = pexConfig.Field(
755 dtype=bool,
756 doc="Use the gain values from the Photon Transfer Curve?",
757 default=False,
758 )
759 normalizeGains = pexConfig.Field(
760 dtype=bool,
761 doc="Normalize all the amplifiers in each CCD to have the same median value.",
762 default=False,
763 )
765 # Fringe correction.
766 doFringe = pexConfig.Field(
767 dtype=bool,
768 doc="Apply fringe correction?",
769 default=True,
770 )
771 fringe = pexConfig.ConfigurableField(
772 target=FringeTask,
773 doc="Fringe subtraction task",
774 )
775 fringeAfterFlat = pexConfig.Field(
776 dtype=bool,
777 doc="Do fringe subtraction after flat-fielding?",
778 default=True,
779 )
781 # Amp offset correction.
782 doAmpOffset = pexConfig.Field(
783 doc="Calculate and apply amp offset corrections?",
784 dtype=bool,
785 default=False,
786 )
787 ampOffset = pexConfig.ConfigurableField(
788 doc="Amp offset correction task.",
789 target=AmpOffsetTask,
790 )
792 # Initial CCD-level background statistics options.
793 doMeasureBackground = pexConfig.Field(
794 dtype=bool,
795 doc="Measure the background level on the reduced image?",
796 default=False,
797 )
799 # Camera-specific masking configuration.
800 doCameraSpecificMasking = pexConfig.Field(
801 dtype=bool,
802 doc="Mask camera-specific bad regions?",
803 default=False,
804 )
805 masking = pexConfig.ConfigurableField(
806 target=MaskingTask,
807 doc="Masking task."
808 )
810 # Interpolation options.
811 doInterpolate = pexConfig.Field(
812 dtype=bool,
813 doc="Interpolate masked pixels?",
814 default=True,
815 )
816 doSaturationInterpolation = pexConfig.Field(
817 dtype=bool,
818 doc="Perform interpolation over pixels masked as saturated?"
819 " NB: This is independent of doSaturation; if that is False this plane"
820 " will likely be blank, resulting in a no-op here.",
821 default=True,
822 )
823 doNanInterpolation = pexConfig.Field(
824 dtype=bool,
825 doc="Perform interpolation over pixels masked as NaN?"
826 " NB: This is independent of doNanMasking; if that is False this plane"
827 " will likely be blank, resulting in a no-op here.",
828 default=True,
829 )
830 doNanInterpAfterFlat = pexConfig.Field(
831 dtype=bool,
832 doc=("If True, ensure we interpolate NaNs after flat-fielding, even if we "
833 "also have to interpolate them before flat-fielding."),
834 default=False,
835 )
836 maskListToInterpolate = pexConfig.ListField(
837 dtype=str,
838 doc="List of mask planes that should be interpolated.",
839 default=['SAT', 'BAD'],
840 )
841 doSaveInterpPixels = pexConfig.Field(
842 dtype=bool,
843 doc="Save a copy of the pre-interpolated pixel values?",
844 default=False,
845 )
847 # Default photometric calibration options.
848 fluxMag0T1 = pexConfig.DictField(
849 keytype=str,
850 itemtype=float,
851 doc="The approximate flux of a zero-magnitude object in a one-second exposure, per filter.",
852 default=dict((f, pow(10.0, 0.4*m)) for f, m in (("Unknown", 28.0),
853 ))
854 )
855 defaultFluxMag0T1 = pexConfig.Field(
856 dtype=float,
857 doc="Default value for fluxMag0T1 (for an unrecognized filter).",
858 default=pow(10.0, 0.4*28.0)
859 )
861 # Vignette correction configuration.
862 doVignette = pexConfig.Field(
863 dtype=bool,
864 doc="Apply vignetting parameters?",
865 default=False,
866 )
867 vignette = pexConfig.ConfigurableField(
868 target=VignetteTask,
869 doc="Vignetting task.",
870 )
872 # Transmission curve configuration.
873 doAttachTransmissionCurve = pexConfig.Field(
874 dtype=bool,
875 default=False,
876 doc="Construct and attach a wavelength-dependent throughput curve for this CCD image?"
877 )
878 doUseOpticsTransmission = pexConfig.Field(
879 dtype=bool,
880 default=True,
881 doc="Load and use transmission_optics (if doAttachTransmissionCurve is True)?"
882 )
883 doUseFilterTransmission = pexConfig.Field(
884 dtype=bool,
885 default=True,
886 doc="Load and use transmission_filter (if doAttachTransmissionCurve is True)?"
887 )
888 doUseSensorTransmission = pexConfig.Field(
889 dtype=bool,
890 default=True,
891 doc="Load and use transmission_sensor (if doAttachTransmissionCurve is True)?"
892 )
893 doUseAtmosphereTransmission = pexConfig.Field(
894 dtype=bool,
895 default=True,
896 doc="Load and use transmission_atmosphere (if doAttachTransmissionCurve is True)?"
897 )
899 # Illumination correction.
900 doIlluminationCorrection = pexConfig.Field(
901 dtype=bool,
902 default=False,
903 doc="Perform illumination correction?"
904 )
905 illuminationCorrectionDataProductName = pexConfig.Field(
906 dtype=str,
907 doc="Name of the illumination correction data product.",
908 default="illumcor",
909 )
910 illumScale = pexConfig.Field(
911 dtype=float,
912 doc="Scale factor for the illumination correction.",
913 default=1.0,
914 )
915 illumFilters = pexConfig.ListField(
916 dtype=str,
917 default=[],
918 doc="Only perform illumination correction for these filters."
919 )
921 # Write the outputs to disk. If ISR is run as a subtask, this may not
922 # be needed.
923 doWrite = pexConfig.Field(
924 dtype=bool,
925 doc="Persist postISRCCD?",
926 default=True,
927 )
929 def validate(self):
930 super().validate()
931 if self.doFlat and self.doApplyGains:
932 raise ValueError("You may not specify both doFlat and doApplyGains")
933 if self.doBiasBeforeOverscan and self.doTrimToMatchCalib:
934 raise ValueError("You may not specify both doBiasBeforeOverscan and doTrimToMatchCalib")
935 if self.doSaturationInterpolation and self.saturatedMaskName not in self.maskListToInterpolate:
936 self.maskListToInterpolate.append(self.saturatedMaskName)
937 if not self.doSaturationInterpolation and self.saturatedMaskName in self.maskListToInterpolate:
938 self.maskListToInterpolate.remove(self.saturatedMaskName)
939 if self.doNanInterpolation and "UNMASKEDNAN" not in self.maskListToInterpolate:
940 self.maskListToInterpolate.append("UNMASKEDNAN")
943class IsrTask(pipeBase.PipelineTask, pipeBase.CmdLineTask):
944 """Apply common instrument signature correction algorithms to a raw frame.
946 The process for correcting imaging data is very similar from
947 camera to camera. This task provides a vanilla implementation of
948 doing these corrections, including the ability to turn certain
949 corrections off if they are not needed. The inputs to the primary
950 method, `run()`, are a raw exposure to be corrected and the
951 calibration data products. The raw input is a single chip sized
952 mosaic of all amps including overscans and other non-science
953 pixels. The method `runDataRef()` identifies and defines the
954 calibration data products, and is intended for use by a
955 `lsst.pipe.base.cmdLineTask.CmdLineTask` and takes as input only a
956 `daf.persistence.butlerSubset.ButlerDataRef`. This task may be
957 subclassed for different camera, although the most camera specific
958 methods have been split into subtasks that can be redirected
959 appropriately.
961 The __init__ method sets up the subtasks for ISR processing, using
962 the defaults from `lsst.ip.isr`.
964 Parameters
965 ----------
966 args : `list`
967 Positional arguments passed to the Task constructor.
968 None used at this time.
969 kwargs : `dict`, optional
970 Keyword arguments passed on to the Task constructor.
971 None used at this time.
972 """
973 ConfigClass = IsrTaskConfig
974 _DefaultName = "isr"
976 def __init__(self, **kwargs):
977 super().__init__(**kwargs)
978 self.makeSubtask("assembleCcd")
979 self.makeSubtask("crosstalk")
980 self.makeSubtask("strayLight")
981 self.makeSubtask("fringe")
982 self.makeSubtask("masking")
983 self.makeSubtask("overscan")
984 self.makeSubtask("vignette")
985 self.makeSubtask("ampOffset")
987 def runQuantum(self, butlerQC, inputRefs, outputRefs):
988 inputs = butlerQC.get(inputRefs)
990 try:
991 inputs['detectorNum'] = inputRefs.ccdExposure.dataId['detector']
992 except Exception as e:
993 raise ValueError("Failure to find valid detectorNum value for Dataset %s: %s." %
994 (inputRefs, e))
996 inputs['isGen3'] = True
998 detector = inputs['ccdExposure'].getDetector()
1000 if self.config.doCrosstalk is True:
1001 # Crosstalk sources need to be defined by the pipeline
1002 # yaml if they exist.
1003 if 'crosstalk' in inputs and inputs['crosstalk'] is not None:
1004 if not isinstance(inputs['crosstalk'], CrosstalkCalib):
1005 inputs['crosstalk'] = CrosstalkCalib.fromTable(inputs['crosstalk'])
1006 else:
1007 coeffVector = (self.config.crosstalk.crosstalkValues
1008 if self.config.crosstalk.useConfigCoefficients else None)
1009 crosstalkCalib = CrosstalkCalib().fromDetector(detector, coeffVector=coeffVector)
1010 inputs['crosstalk'] = crosstalkCalib
1011 if inputs['crosstalk'].interChip and len(inputs['crosstalk'].interChip) > 0:
1012 if 'crosstalkSources' not in inputs:
1013 self.log.warning("No crosstalkSources found for chip with interChip terms!")
1015 if self.doLinearize(detector) is True:
1016 if 'linearizer' in inputs:
1017 if isinstance(inputs['linearizer'], dict):
1018 linearizer = linearize.Linearizer(detector=detector, log=self.log)
1019 linearizer.fromYaml(inputs['linearizer'])
1020 self.log.warning("Dictionary linearizers will be deprecated in DM-28741.")
1021 elif isinstance(inputs['linearizer'], numpy.ndarray):
1022 linearizer = linearize.Linearizer(table=inputs.get('linearizer', None),
1023 detector=detector,
1024 log=self.log)
1025 self.log.warning("Bare lookup table linearizers will be deprecated in DM-28741.")
1026 else:
1027 linearizer = inputs['linearizer']
1028 linearizer.log = self.log
1029 inputs['linearizer'] = linearizer
1030 else:
1031 inputs['linearizer'] = linearize.Linearizer(detector=detector, log=self.log)
1032 self.log.warning("Constructing linearizer from cameraGeom information.")
1034 if self.config.doDefect is True:
1035 if "defects" in inputs and inputs['defects'] is not None:
1036 # defects is loaded as a BaseCatalog with columns
1037 # x0, y0, width, height. Masking expects a list of defects
1038 # defined by their bounding box
1039 if not isinstance(inputs["defects"], Defects):
1040 inputs["defects"] = Defects.fromTable(inputs["defects"])
1042 # Load the correct style of brighter-fatter kernel, and repack
1043 # the information as a numpy array.
1044 if self.config.doBrighterFatter:
1045 brighterFatterKernel = inputs.pop('newBFKernel', None)
1046 if brighterFatterKernel is None:
1047 brighterFatterKernel = inputs.get('bfKernel', None)
1049 if brighterFatterKernel is not None and not isinstance(brighterFatterKernel, numpy.ndarray):
1050 # This is a ISR calib kernel
1051 detName = detector.getName()
1052 level = brighterFatterKernel.level
1054 # This is expected to be a dictionary of amp-wise gains.
1055 inputs['bfGains'] = brighterFatterKernel.gain
1056 if self.config.brighterFatterLevel == 'DETECTOR':
1057 if level == 'DETECTOR':
1058 if detName in brighterFatterKernel.detKernels:
1059 inputs['bfKernel'] = brighterFatterKernel.detKernels[detName]
1060 else:
1061 raise RuntimeError("Failed to extract kernel from new-style BF kernel.")
1062 elif level == 'AMP':
1063 self.log.warning("Making DETECTOR level kernel from AMP based brighter "
1064 "fatter kernels.")
1065 brighterFatterKernel.makeDetectorKernelFromAmpwiseKernels(detName)
1066 inputs['bfKernel'] = brighterFatterKernel.detKernels[detName]
1067 elif self.config.brighterFatterLevel == 'AMP':
1068 raise NotImplementedError("Per-amplifier brighter-fatter correction not implemented")
1070 if self.config.doFringe is True and self.fringe.checkFilter(inputs['ccdExposure']):
1071 expId = inputs['ccdExposure'].getInfo().getVisitInfo().getExposureId()
1072 inputs['fringes'] = self.fringe.loadFringes(inputs['fringes'],
1073 expId=expId,
1074 assembler=self.assembleCcd
1075 if self.config.doAssembleIsrExposures else None)
1076 else:
1077 inputs['fringes'] = pipeBase.Struct(fringes=None)
1079 if self.config.doStrayLight is True and self.strayLight.checkFilter(inputs['ccdExposure']):
1080 if 'strayLightData' not in inputs:
1081 inputs['strayLightData'] = None
1083 outputs = self.run(**inputs)
1084 butlerQC.put(outputs, outputRefs)
1086 def readIsrData(self, dataRef, rawExposure):
1087 """Retrieve necessary frames for instrument signature removal.
1089 Pre-fetching all required ISR data products limits the IO
1090 required by the ISR. Any conflict between the calibration data
1091 available and that needed for ISR is also detected prior to
1092 doing processing, allowing it to fail quickly.
1094 Parameters
1095 ----------
1096 dataRef : `daf.persistence.butlerSubset.ButlerDataRef`
1097 Butler reference of the detector data to be processed
1098 rawExposure : `afw.image.Exposure`
1099 The raw exposure that will later be corrected with the
1100 retrieved calibration data; should not be modified in this
1101 method.
1103 Returns
1104 -------
1105 result : `lsst.pipe.base.Struct`
1106 Result struct with components (which may be `None`):
1107 - ``bias``: bias calibration frame (`afw.image.Exposure`)
1108 - ``linearizer``: functor for linearization
1109 (`ip.isr.linearize.LinearizeBase`)
1110 - ``crosstalkSources``: list of possible crosstalk sources (`list`)
1111 - ``dark``: dark calibration frame (`afw.image.Exposure`)
1112 - ``flat``: flat calibration frame (`afw.image.Exposure`)
1113 - ``bfKernel``: Brighter-Fatter kernel (`numpy.ndarray`)
1114 - ``defects``: list of defects (`lsst.ip.isr.Defects`)
1115 - ``fringes``: `lsst.pipe.base.Struct` with components:
1116 - ``fringes``: fringe calibration frame (`afw.image.Exposure`)
1117 - ``seed``: random seed derived from the ccdExposureId for random
1118 number generator (`uint32`).
1119 - ``opticsTransmission``: `lsst.afw.image.TransmissionCurve`
1120 A ``TransmissionCurve`` that represents the throughput of the
1121 optics, to be evaluated in focal-plane coordinates.
1122 - ``filterTransmission`` : `lsst.afw.image.TransmissionCurve`
1123 A ``TransmissionCurve`` that represents the throughput of the
1124 filter itself, to be evaluated in focal-plane coordinates.
1125 - ``sensorTransmission`` : `lsst.afw.image.TransmissionCurve`
1126 A ``TransmissionCurve`` that represents the throughput of the
1127 sensor itself, to be evaluated in post-assembly trimmed
1128 detector coordinates.
1129 - ``atmosphereTransmission`` : `lsst.afw.image.TransmissionCurve`
1130 A ``TransmissionCurve`` that represents the throughput of the
1131 atmosphere, assumed to be spatially constant.
1132 - ``strayLightData`` : `object`
1133 An opaque object containing calibration information for
1134 stray-light correction. If `None`, no correction will be
1135 performed.
1136 - ``illumMaskedImage`` : illumination correction image
1137 (`lsst.afw.image.MaskedImage`)
1139 Raises
1140 ------
1141 NotImplementedError :
1142 Raised if a per-amplifier brighter-fatter kernel is requested by
1143 the configuration.
1144 """
1145 try:
1146 dateObs = rawExposure.getInfo().getVisitInfo().getDate()
1147 dateObs = dateObs.toPython().isoformat()
1148 except RuntimeError:
1149 self.log.warning("Unable to identify dateObs for rawExposure.")
1150 dateObs = None
1152 ccd = rawExposure.getDetector()
1153 filterLabel = rawExposure.getFilterLabel()
1154 physicalFilter = isrFunctions.getPhysicalFilter(filterLabel, self.log)
1155 rawExposure.mask.addMaskPlane("UNMASKEDNAN") # needed to match pre DM-15862 processing.
1156 biasExposure = (self.getIsrExposure(dataRef, self.config.biasDataProductName)
1157 if self.config.doBias else None)
1158 # immediate=True required for functors and linearizers are functors
1159 # see ticket DM-6515
1160 linearizer = (dataRef.get("linearizer", immediate=True)
1161 if self.doLinearize(ccd) else None)
1162 if linearizer is not None and not isinstance(linearizer, numpy.ndarray):
1163 linearizer.log = self.log
1164 if isinstance(linearizer, numpy.ndarray):
1165 linearizer = linearize.Linearizer(table=linearizer, detector=ccd)
1167 crosstalkCalib = None
1168 if self.config.doCrosstalk:
1169 try:
1170 crosstalkCalib = dataRef.get("crosstalk", immediate=True)
1171 except NoResults:
1172 coeffVector = (self.config.crosstalk.crosstalkValues
1173 if self.config.crosstalk.useConfigCoefficients else None)
1174 crosstalkCalib = CrosstalkCalib().fromDetector(ccd, coeffVector=coeffVector)
1175 crosstalkSources = (self.crosstalk.prepCrosstalk(dataRef, crosstalkCalib)
1176 if self.config.doCrosstalk else None)
1178 darkExposure = (self.getIsrExposure(dataRef, self.config.darkDataProductName)
1179 if self.config.doDark else None)
1180 flatExposure = (self.getIsrExposure(dataRef, self.config.flatDataProductName,
1181 dateObs=dateObs)
1182 if self.config.doFlat else None)
1184 brighterFatterKernel = None
1185 brighterFatterGains = None
1186 if self.config.doBrighterFatter is True:
1187 try:
1188 # Use the new-style cp_pipe version of the kernel if it exists
1189 # If using a new-style kernel, always use the self-consistent
1190 # gains, i.e. the ones inside the kernel object itself
1191 brighterFatterKernel = dataRef.get("brighterFatterKernel")
1192 brighterFatterGains = brighterFatterKernel.gain
1193 self.log.info("New style brighter-fatter kernel (brighterFatterKernel) loaded")
1194 except NoResults:
1195 try: # Fall back to the old-style numpy-ndarray style kernel if necessary.
1196 brighterFatterKernel = dataRef.get("bfKernel")
1197 self.log.info("Old style brighter-fatter kernel (bfKernel) loaded")
1198 except NoResults:
1199 brighterFatterKernel = None
1200 if brighterFatterKernel is not None and not isinstance(brighterFatterKernel, numpy.ndarray):
1201 # If the kernel is not an ndarray, it's the cp_pipe version
1202 # so extract the kernel for this detector, or raise an error
1203 if self.config.brighterFatterLevel == 'DETECTOR':
1204 if brighterFatterKernel.detKernels:
1205 brighterFatterKernel = brighterFatterKernel.detKernels[ccd.getName()]
1206 else:
1207 raise RuntimeError("Failed to extract kernel from new-style BF kernel.")
1208 else:
1209 # TODO DM-15631 for implementing this
1210 raise NotImplementedError("Per-amplifier brighter-fatter correction not implemented")
1212 defectList = (dataRef.get("defects")
1213 if self.config.doDefect else None)
1214 expId = rawExposure.getInfo().getVisitInfo().getExposureId()
1215 fringeStruct = (self.fringe.readFringes(dataRef, expId=expId, assembler=self.assembleCcd
1216 if self.config.doAssembleIsrExposures else None)
1217 if self.config.doFringe and self.fringe.checkFilter(rawExposure)
1218 else pipeBase.Struct(fringes=None))
1220 if self.config.doAttachTransmissionCurve:
1221 opticsTransmission = (dataRef.get("transmission_optics")
1222 if self.config.doUseOpticsTransmission else None)
1223 filterTransmission = (dataRef.get("transmission_filter")
1224 if self.config.doUseFilterTransmission else None)
1225 sensorTransmission = (dataRef.get("transmission_sensor")
1226 if self.config.doUseSensorTransmission else None)
1227 atmosphereTransmission = (dataRef.get("transmission_atmosphere")
1228 if self.config.doUseAtmosphereTransmission else None)
1229 else:
1230 opticsTransmission = None
1231 filterTransmission = None
1232 sensorTransmission = None
1233 atmosphereTransmission = None
1235 if self.config.doStrayLight:
1236 strayLightData = self.strayLight.readIsrData(dataRef, rawExposure)
1237 else:
1238 strayLightData = None
1240 illumMaskedImage = (self.getIsrExposure(dataRef,
1241 self.config.illuminationCorrectionDataProductName).getMaskedImage()
1242 if (self.config.doIlluminationCorrection
1243 and physicalFilter in self.config.illumFilters)
1244 else None)
1246 # Struct should include only kwargs to run()
1247 return pipeBase.Struct(bias=biasExposure,
1248 linearizer=linearizer,
1249 crosstalk=crosstalkCalib,
1250 crosstalkSources=crosstalkSources,
1251 dark=darkExposure,
1252 flat=flatExposure,
1253 bfKernel=brighterFatterKernel,
1254 bfGains=brighterFatterGains,
1255 defects=defectList,
1256 fringes=fringeStruct,
1257 opticsTransmission=opticsTransmission,
1258 filterTransmission=filterTransmission,
1259 sensorTransmission=sensorTransmission,
1260 atmosphereTransmission=atmosphereTransmission,
1261 strayLightData=strayLightData,
1262 illumMaskedImage=illumMaskedImage
1263 )
1265 @pipeBase.timeMethod
1266 def run(self, ccdExposure, *, camera=None, bias=None, linearizer=None,
1267 crosstalk=None, crosstalkSources=None,
1268 dark=None, flat=None, ptc=None, bfKernel=None, bfGains=None, defects=None,
1269 fringes=pipeBase.Struct(fringes=None), opticsTransmission=None, filterTransmission=None,
1270 sensorTransmission=None, atmosphereTransmission=None,
1271 detectorNum=None, strayLightData=None, illumMaskedImage=None,
1272 isGen3=False,
1273 ):
1274 """Perform instrument signature removal on an exposure.
1276 Steps included in the ISR processing, in order performed, are:
1277 - saturation and suspect pixel masking
1278 - overscan subtraction
1279 - CCD assembly of individual amplifiers
1280 - bias subtraction
1281 - variance image construction
1282 - linearization of non-linear response
1283 - crosstalk masking
1284 - brighter-fatter correction
1285 - dark subtraction
1286 - fringe correction
1287 - stray light subtraction
1288 - flat correction
1289 - masking of known defects and camera specific features
1290 - vignette calculation
1291 - appending transmission curve and distortion model
1293 Parameters
1294 ----------
1295 ccdExposure : `lsst.afw.image.Exposure`
1296 The raw exposure that is to be run through ISR. The
1297 exposure is modified by this method.
1298 camera : `lsst.afw.cameraGeom.Camera`, optional
1299 The camera geometry for this exposure. Required if
1300 one or more of ``ccdExposure``, ``bias``, ``dark``, or
1301 ``flat`` does not have an associated detector.
1302 bias : `lsst.afw.image.Exposure`, optional
1303 Bias calibration frame.
1304 linearizer : `lsst.ip.isr.linearize.LinearizeBase`, optional
1305 Functor for linearization.
1306 crosstalk : `lsst.ip.isr.crosstalk.CrosstalkCalib`, optional
1307 Calibration for crosstalk.
1308 crosstalkSources : `list`, optional
1309 List of possible crosstalk sources.
1310 dark : `lsst.afw.image.Exposure`, optional
1311 Dark calibration frame.
1312 flat : `lsst.afw.image.Exposure`, optional
1313 Flat calibration frame.
1314 ptc : `lsst.ip.isr.PhotonTransferCurveDataset`, optional
1315 Photon transfer curve dataset, with, e.g., gains
1316 and read noise.
1317 bfKernel : `numpy.ndarray`, optional
1318 Brighter-fatter kernel.
1319 bfGains : `dict` of `float`, optional
1320 Gains used to override the detector's nominal gains for the
1321 brighter-fatter correction. A dict keyed by amplifier name for
1322 the detector in question.
1323 defects : `lsst.ip.isr.Defects`, optional
1324 List of defects.
1325 fringes : `lsst.pipe.base.Struct`, optional
1326 Struct containing the fringe correction data, with
1327 elements:
1328 - ``fringes``: fringe calibration frame (`afw.image.Exposure`)
1329 - ``seed``: random seed derived from the ccdExposureId for random
1330 number generator (`uint32`)
1331 opticsTransmission: `lsst.afw.image.TransmissionCurve`, optional
1332 A ``TransmissionCurve`` that represents the throughput of the,
1333 optics, to be evaluated in focal-plane coordinates.
1334 filterTransmission : `lsst.afw.image.TransmissionCurve`
1335 A ``TransmissionCurve`` that represents the throughput of the
1336 filter itself, to be evaluated in focal-plane coordinates.
1337 sensorTransmission : `lsst.afw.image.TransmissionCurve`
1338 A ``TransmissionCurve`` that represents the throughput of the
1339 sensor itself, to be evaluated in post-assembly trimmed detector
1340 coordinates.
1341 atmosphereTransmission : `lsst.afw.image.TransmissionCurve`
1342 A ``TransmissionCurve`` that represents the throughput of the
1343 atmosphere, assumed to be spatially constant.
1344 detectorNum : `int`, optional
1345 The integer number for the detector to process.
1346 isGen3 : bool, optional
1347 Flag this call to run() as using the Gen3 butler environment.
1348 strayLightData : `object`, optional
1349 Opaque object containing calibration information for stray-light
1350 correction. If `None`, no correction will be performed.
1351 illumMaskedImage : `lsst.afw.image.MaskedImage`, optional
1352 Illumination correction image.
1354 Returns
1355 -------
1356 result : `lsst.pipe.base.Struct`
1357 Result struct with component:
1358 - ``exposure`` : `afw.image.Exposure`
1359 The fully ISR corrected exposure.
1360 - ``outputExposure`` : `afw.image.Exposure`
1361 An alias for `exposure`
1362 - ``ossThumb`` : `numpy.ndarray`
1363 Thumbnail image of the exposure after overscan subtraction.
1364 - ``flattenedThumb`` : `numpy.ndarray`
1365 Thumbnail image of the exposure after flat-field correction.
1367 Raises
1368 ------
1369 RuntimeError
1370 Raised if a configuration option is set to True, but the
1371 required calibration data has not been specified.
1373 Notes
1374 -----
1375 The current processed exposure can be viewed by setting the
1376 appropriate lsstDebug entries in the `debug.display`
1377 dictionary. The names of these entries correspond to some of
1378 the IsrTaskConfig Boolean options, with the value denoting the
1379 frame to use. The exposure is shown inside the matching
1380 option check and after the processing of that step has
1381 finished. The steps with debug points are:
1383 doAssembleCcd
1384 doBias
1385 doCrosstalk
1386 doBrighterFatter
1387 doDark
1388 doFringe
1389 doStrayLight
1390 doFlat
1392 In addition, setting the "postISRCCD" entry displays the
1393 exposure after all ISR processing has finished.
1395 """
1397 if isGen3 is True:
1398 # Gen3 currently cannot automatically do configuration overrides.
1399 # DM-15257 looks to discuss this issue.
1400 # Configure input exposures;
1402 ccdExposure = self.ensureExposure(ccdExposure, camera, detectorNum)
1403 bias = self.ensureExposure(bias, camera, detectorNum)
1404 dark = self.ensureExposure(dark, camera, detectorNum)
1405 flat = self.ensureExposure(flat, camera, detectorNum)
1406 else:
1407 if isinstance(ccdExposure, ButlerDataRef):
1408 return self.runDataRef(ccdExposure)
1410 ccd = ccdExposure.getDetector()
1411 filterLabel = ccdExposure.getFilterLabel()
1412 physicalFilter = isrFunctions.getPhysicalFilter(filterLabel, self.log)
1414 if not ccd:
1415 assert not self.config.doAssembleCcd, "You need a Detector to run assembleCcd."
1416 ccd = [FakeAmp(ccdExposure, self.config)]
1418 # Validate Input
1419 if self.config.doBias and bias is None:
1420 raise RuntimeError("Must supply a bias exposure if config.doBias=True.")
1421 if self.doLinearize(ccd) and linearizer is None:
1422 raise RuntimeError("Must supply a linearizer if config.doLinearize=True for this detector.")
1423 if self.config.doBrighterFatter and bfKernel is None:
1424 raise RuntimeError("Must supply a kernel if config.doBrighterFatter=True.")
1425 if self.config.doDark and dark is None:
1426 raise RuntimeError("Must supply a dark exposure if config.doDark=True.")
1427 if self.config.doFlat and flat is None:
1428 raise RuntimeError("Must supply a flat exposure if config.doFlat=True.")
1429 if self.config.doDefect and defects is None:
1430 raise RuntimeError("Must supply defects if config.doDefect=True.")
1431 if (self.config.doFringe and physicalFilter in self.fringe.config.filters
1432 and fringes.fringes is None):
1433 # The `fringes` object needs to be a pipeBase.Struct, as
1434 # we use it as a `dict` for the parameters of
1435 # `FringeTask.run()`. The `fringes.fringes` `list` may
1436 # not be `None` if `doFringe=True`. Otherwise, raise.
1437 raise RuntimeError("Must supply fringe exposure as a pipeBase.Struct.")
1438 if (self.config.doIlluminationCorrection and physicalFilter in self.config.illumFilters
1439 and illumMaskedImage is None):
1440 raise RuntimeError("Must supply an illumcor if config.doIlluminationCorrection=True.")
1442 # Begin ISR processing.
1443 if self.config.doConvertIntToFloat:
1444 self.log.info("Converting exposure to floating point values.")
1445 ccdExposure = self.convertIntToFloat(ccdExposure)
1447 if self.config.doBias and self.config.doBiasBeforeOverscan:
1448 self.log.info("Applying bias correction.")
1449 isrFunctions.biasCorrection(ccdExposure.getMaskedImage(), bias.getMaskedImage(),
1450 trimToFit=self.config.doTrimToMatchCalib)
1451 self.debugView(ccdExposure, "doBias")
1453 # Amplifier level processing.
1454 overscans = []
1455 for amp in ccd:
1456 # if ccdExposure is one amp,
1457 # check for coverage to prevent performing ops multiple times
1458 if ccdExposure.getBBox().contains(amp.getBBox()):
1459 # Check for fully masked bad amplifiers,
1460 # and generate masks for SUSPECT and SATURATED values.
1461 badAmp = self.maskAmplifier(ccdExposure, amp, defects)
1463 if self.config.doOverscan and not badAmp:
1464 # Overscan correction on amp-by-amp basis.
1465 overscanResults = self.overscanCorrection(ccdExposure, amp)
1466 self.log.debug("Corrected overscan for amplifier %s.", amp.getName())
1467 if overscanResults is not None and \
1468 self.config.qa is not None and self.config.qa.saveStats is True:
1469 if isinstance(overscanResults.overscanFit, float):
1470 qaMedian = overscanResults.overscanFit
1471 qaStdev = float("NaN")
1472 else:
1473 qaStats = afwMath.makeStatistics(overscanResults.overscanFit,
1474 afwMath.MEDIAN | afwMath.STDEVCLIP)
1475 qaMedian = qaStats.getValue(afwMath.MEDIAN)
1476 qaStdev = qaStats.getValue(afwMath.STDEVCLIP)
1478 self.metadata.set(f"FIT MEDIAN {amp.getName()}", qaMedian)
1479 self.metadata.set(f"FIT STDEV {amp.getName()}", qaStdev)
1480 self.log.debug(" Overscan stats for amplifer %s: %f +/- %f",
1481 amp.getName(), qaMedian, qaStdev)
1483 # Residuals after overscan correction
1484 qaStatsAfter = afwMath.makeStatistics(overscanResults.overscanImage,
1485 afwMath.MEDIAN | afwMath.STDEVCLIP)
1486 qaMedianAfter = qaStatsAfter.getValue(afwMath.MEDIAN)
1487 qaStdevAfter = qaStatsAfter.getValue(afwMath.STDEVCLIP)
1489 self.metadata.set(f"RESIDUAL MEDIAN {amp.getName()}", qaMedianAfter)
1490 self.metadata.set(f"RESIDUAL STDEV {amp.getName()}", qaStdevAfter)
1491 self.log.debug(" Overscan stats for amplifer %s after correction: %f +/- %f",
1492 amp.getName(), qaMedianAfter, qaStdevAfter)
1494 ccdExposure.getMetadata().set('OVERSCAN', "Overscan corrected")
1495 else:
1496 if badAmp:
1497 self.log.warning("Amplifier %s is bad.", amp.getName())
1498 overscanResults = None
1500 overscans.append(overscanResults if overscanResults is not None else None)
1501 else:
1502 self.log.info("Skipped OSCAN for %s.", amp.getName())
1504 if self.config.doCrosstalk and self.config.doCrosstalkBeforeAssemble:
1505 self.log.info("Applying crosstalk correction.")
1506 self.crosstalk.run(ccdExposure, crosstalk=crosstalk,
1507 crosstalkSources=crosstalkSources, camera=camera)
1508 self.debugView(ccdExposure, "doCrosstalk")
1510 if self.config.doAssembleCcd:
1511 self.log.info("Assembling CCD from amplifiers.")
1512 ccdExposure = self.assembleCcd.assembleCcd(ccdExposure)
1514 if self.config.expectWcs and not ccdExposure.getWcs():
1515 self.log.warning("No WCS found in input exposure.")
1516 self.debugView(ccdExposure, "doAssembleCcd")
1518 ossThumb = None
1519 if self.config.qa.doThumbnailOss:
1520 ossThumb = isrQa.makeThumbnail(ccdExposure, isrQaConfig=self.config.qa)
1522 if self.config.doBias and not self.config.doBiasBeforeOverscan:
1523 self.log.info("Applying bias correction.")
1524 isrFunctions.biasCorrection(ccdExposure.getMaskedImage(), bias.getMaskedImage(),
1525 trimToFit=self.config.doTrimToMatchCalib)
1526 self.debugView(ccdExposure, "doBias")
1528 if self.config.doVariance:
1529 for amp, overscanResults in zip(ccd, overscans):
1530 if ccdExposure.getBBox().contains(amp.getBBox()):
1531 self.log.debug("Constructing variance map for amplifer %s.", amp.getName())
1532 ampExposure = ccdExposure.Factory(ccdExposure, amp.getBBox())
1533 if overscanResults is not None:
1534 self.updateVariance(ampExposure, amp,
1535 overscanImage=overscanResults.overscanImage,
1536 ptcDataset=ptc)
1537 else:
1538 self.updateVariance(ampExposure, amp,
1539 overscanImage=None,
1540 ptcDataset=ptc)
1541 if self.config.qa is not None and self.config.qa.saveStats is True:
1542 qaStats = afwMath.makeStatistics(ampExposure.getVariance(),
1543 afwMath.MEDIAN | afwMath.STDEVCLIP)
1544 self.metadata.set(f"ISR VARIANCE {amp.getName()} MEDIAN",
1545 qaStats.getValue(afwMath.MEDIAN))
1546 self.metadata.set(f"ISR VARIANCE {amp.getName()} STDEV",
1547 qaStats.getValue(afwMath.STDEVCLIP))
1548 self.log.debug(" Variance stats for amplifer %s: %f +/- %f.",
1549 amp.getName(), qaStats.getValue(afwMath.MEDIAN),
1550 qaStats.getValue(afwMath.STDEVCLIP))
1551 if self.config.maskNegativeVariance:
1552 self.maskNegativeVariance(ccdExposure)
1554 if self.doLinearize(ccd):
1555 self.log.info("Applying linearizer.")
1556 linearizer.applyLinearity(image=ccdExposure.getMaskedImage().getImage(),
1557 detector=ccd, log=self.log)
1559 if self.config.doCrosstalk and not self.config.doCrosstalkBeforeAssemble:
1560 self.log.info("Applying crosstalk correction.")
1561 self.crosstalk.run(ccdExposure, crosstalk=crosstalk,
1562 crosstalkSources=crosstalkSources, isTrimmed=True)
1563 self.debugView(ccdExposure, "doCrosstalk")
1565 # Masking block. Optionally mask known defects, NAN/inf pixels,
1566 # widen trails, and do anything else the camera needs. Saturated and
1567 # suspect pixels have already been masked.
1568 if self.config.doDefect:
1569 self.log.info("Masking defects.")
1570 self.maskDefect(ccdExposure, defects)
1572 if self.config.numEdgeSuspect > 0:
1573 self.log.info("Masking edges as SUSPECT.")
1574 self.maskEdges(ccdExposure, numEdgePixels=self.config.numEdgeSuspect,
1575 maskPlane="SUSPECT", level=self.config.edgeMaskLevel)
1577 if self.config.doNanMasking:
1578 self.log.info("Masking non-finite (NAN, inf) value pixels.")
1579 self.maskNan(ccdExposure)
1581 if self.config.doWidenSaturationTrails:
1582 self.log.info("Widening saturation trails.")
1583 isrFunctions.widenSaturationTrails(ccdExposure.getMaskedImage().getMask())
1585 if self.config.doCameraSpecificMasking:
1586 self.log.info("Masking regions for camera specific reasons.")
1587 self.masking.run(ccdExposure)
1589 if self.config.doBrighterFatter:
1590 # We need to apply flats and darks before we can interpolate, and
1591 # we need to interpolate before we do B-F, but we do B-F without
1592 # the flats and darks applied so we can work in units of electrons
1593 # or holes. This context manager applies and then removes the darks
1594 # and flats.
1595 #
1596 # We also do not want to interpolate values here, so operate on
1597 # temporary images so we can apply only the BF-correction and roll
1598 # back the interpolation.
1599 interpExp = ccdExposure.clone()
1600 with self.flatContext(interpExp, flat, dark):
1601 isrFunctions.interpolateFromMask(
1602 maskedImage=interpExp.getMaskedImage(),
1603 fwhm=self.config.fwhm,
1604 growSaturatedFootprints=self.config.growSaturationFootprintSize,
1605 maskNameList=list(self.config.brighterFatterMaskListToInterpolate)
1606 )
1607 bfExp = interpExp.clone()
1609 self.log.info("Applying brighter-fatter correction using kernel type %s / gains %s.",
1610 type(bfKernel), type(bfGains))
1611 bfResults = isrFunctions.brighterFatterCorrection(bfExp, bfKernel,
1612 self.config.brighterFatterMaxIter,
1613 self.config.brighterFatterThreshold,
1614 self.config.brighterFatterApplyGain,
1615 bfGains)
1616 if bfResults[1] == self.config.brighterFatterMaxIter:
1617 self.log.warning("Brighter-fatter correction did not converge, final difference %f.",
1618 bfResults[0])
1619 else:
1620 self.log.info("Finished brighter-fatter correction in %d iterations.",
1621 bfResults[1])
1622 image = ccdExposure.getMaskedImage().getImage()
1623 bfCorr = bfExp.getMaskedImage().getImage()
1624 bfCorr -= interpExp.getMaskedImage().getImage()
1625 image += bfCorr
1627 # Applying the brighter-fatter correction applies a
1628 # convolution to the science image. At the edges this
1629 # convolution may not have sufficient valid pixels to
1630 # produce a valid correction. Mark pixels within the size
1631 # of the brighter-fatter kernel as EDGE to warn of this
1632 # fact.
1633 self.log.info("Ensuring image edges are masked as EDGE to the brighter-fatter kernel size.")
1634 self.maskEdges(ccdExposure, numEdgePixels=numpy.max(bfKernel.shape) // 2,
1635 maskPlane="EDGE")
1637 if self.config.brighterFatterMaskGrowSize > 0:
1638 self.log.info("Growing masks to account for brighter-fatter kernel convolution.")
1639 for maskPlane in self.config.brighterFatterMaskListToInterpolate:
1640 isrFunctions.growMasks(ccdExposure.getMask(),
1641 radius=self.config.brighterFatterMaskGrowSize,
1642 maskNameList=maskPlane,
1643 maskValue=maskPlane)
1645 self.debugView(ccdExposure, "doBrighterFatter")
1647 if self.config.doDark:
1648 self.log.info("Applying dark correction.")
1649 self.darkCorrection(ccdExposure, dark)
1650 self.debugView(ccdExposure, "doDark")
1652 if self.config.doFringe and not self.config.fringeAfterFlat:
1653 self.log.info("Applying fringe correction before flat.")
1654 self.fringe.run(ccdExposure, **fringes.getDict())
1655 self.debugView(ccdExposure, "doFringe")
1657 if self.config.doStrayLight and self.strayLight.check(ccdExposure):
1658 self.log.info("Checking strayLight correction.")
1659 self.strayLight.run(ccdExposure, strayLightData)
1660 self.debugView(ccdExposure, "doStrayLight")
1662 if self.config.doFlat:
1663 self.log.info("Applying flat correction.")
1664 self.flatCorrection(ccdExposure, flat)
1665 self.debugView(ccdExposure, "doFlat")
1667 if self.config.doApplyGains:
1668 self.log.info("Applying gain correction instead of flat.")
1669 if self.config.usePtcGains:
1670 self.log.info("Using gains from the Photon Transfer Curve.")
1671 isrFunctions.applyGains(ccdExposure, self.config.normalizeGains,
1672 ptcGains=ptc.gain)
1673 else:
1674 isrFunctions.applyGains(ccdExposure, self.config.normalizeGains)
1676 if self.config.doFringe and self.config.fringeAfterFlat:
1677 self.log.info("Applying fringe correction after flat.")
1678 self.fringe.run(ccdExposure, **fringes.getDict())
1680 if self.config.doVignette:
1681 self.log.info("Constructing Vignette polygon.")
1682 self.vignettePolygon = self.vignette.run(ccdExposure)
1684 if self.config.vignette.doWriteVignettePolygon:
1685 self.setValidPolygonIntersect(ccdExposure, self.vignettePolygon)
1687 if self.config.doAttachTransmissionCurve:
1688 self.log.info("Adding transmission curves.")
1689 isrFunctions.attachTransmissionCurve(ccdExposure, opticsTransmission=opticsTransmission,
1690 filterTransmission=filterTransmission,
1691 sensorTransmission=sensorTransmission,
1692 atmosphereTransmission=atmosphereTransmission)
1694 flattenedThumb = None
1695 if self.config.qa.doThumbnailFlattened:
1696 flattenedThumb = isrQa.makeThumbnail(ccdExposure, isrQaConfig=self.config.qa)
1698 if self.config.doIlluminationCorrection and physicalFilter in self.config.illumFilters:
1699 self.log.info("Performing illumination correction.")
1700 isrFunctions.illuminationCorrection(ccdExposure.getMaskedImage(),
1701 illumMaskedImage, illumScale=self.config.illumScale,
1702 trimToFit=self.config.doTrimToMatchCalib)
1704 preInterpExp = None
1705 if self.config.doSaveInterpPixels:
1706 preInterpExp = ccdExposure.clone()
1708 # Reset and interpolate bad pixels.
1709 #
1710 # Large contiguous bad regions (which should have the BAD mask
1711 # bit set) should have their values set to the image median.
1712 # This group should include defects and bad amplifiers. As the
1713 # area covered by these defects are large, there's little
1714 # reason to expect that interpolation would provide a more
1715 # useful value.
1716 #
1717 # Smaller defects can be safely interpolated after the larger
1718 # regions have had their pixel values reset. This ensures
1719 # that the remaining defects adjacent to bad amplifiers (as an
1720 # example) do not attempt to interpolate extreme values.
1721 if self.config.doSetBadRegions:
1722 badPixelCount, badPixelValue = isrFunctions.setBadRegions(ccdExposure)
1723 if badPixelCount > 0:
1724 self.log.info("Set %d BAD pixels to %f.", badPixelCount, badPixelValue)
1726 if self.config.doInterpolate:
1727 self.log.info("Interpolating masked pixels.")
1728 isrFunctions.interpolateFromMask(
1729 maskedImage=ccdExposure.getMaskedImage(),
1730 fwhm=self.config.fwhm,
1731 growSaturatedFootprints=self.config.growSaturationFootprintSize,
1732 maskNameList=list(self.config.maskListToInterpolate)
1733 )
1735 self.roughZeroPoint(ccdExposure)
1737 # correct for amp offsets within the CCD
1738 if self.config.doAmpOffset:
1739 self.log.info("Correcting amp offsets.")
1740 self.ampOffset.run(ccdExposure)
1742 if self.config.doMeasureBackground:
1743 self.log.info("Measuring background level.")
1744 self.measureBackground(ccdExposure, self.config.qa)
1746 if self.config.qa is not None and self.config.qa.saveStats is True:
1747 for amp in ccd:
1748 ampExposure = ccdExposure.Factory(ccdExposure, amp.getBBox())
1749 qaStats = afwMath.makeStatistics(ampExposure.getImage(),
1750 afwMath.MEDIAN | afwMath.STDEVCLIP)
1751 self.metadata.set("ISR BACKGROUND {} MEDIAN".format(amp.getName()),
1752 qaStats.getValue(afwMath.MEDIAN))
1753 self.metadata.set("ISR BACKGROUND {} STDEV".format(amp.getName()),
1754 qaStats.getValue(afwMath.STDEVCLIP))
1755 self.log.debug(" Background stats for amplifer %s: %f +/- %f",
1756 amp.getName(), qaStats.getValue(afwMath.MEDIAN),
1757 qaStats.getValue(afwMath.STDEVCLIP))
1759 self.debugView(ccdExposure, "postISRCCD")
1761 return pipeBase.Struct(
1762 exposure=ccdExposure,
1763 ossThumb=ossThumb,
1764 flattenedThumb=flattenedThumb,
1766 preInterpExposure=preInterpExp,
1767 outputExposure=ccdExposure,
1768 outputOssThumbnail=ossThumb,
1769 outputFlattenedThumbnail=flattenedThumb,
1770 )
1772 @pipeBase.timeMethod
1773 def runDataRef(self, sensorRef):
1774 """Perform instrument signature removal on a ButlerDataRef of a Sensor.
1776 This method contains the `CmdLineTask` interface to the ISR
1777 processing. All IO is handled here, freeing the `run()` method
1778 to manage only pixel-level calculations. The steps performed
1779 are:
1780 - Read in necessary detrending/isr/calibration data.
1781 - Process raw exposure in `run()`.
1782 - Persist the ISR-corrected exposure as "postISRCCD" if
1783 config.doWrite=True.
1785 Parameters
1786 ----------
1787 sensorRef : `daf.persistence.butlerSubset.ButlerDataRef`
1788 DataRef of the detector data to be processed
1790 Returns
1791 -------
1792 result : `lsst.pipe.base.Struct`
1793 Result struct with component:
1794 - ``exposure`` : `afw.image.Exposure`
1795 The fully ISR corrected exposure.
1797 Raises
1798 ------
1799 RuntimeError
1800 Raised if a configuration option is set to True, but the
1801 required calibration data does not exist.
1803 """
1804 self.log.info("Performing ISR on sensor %s.", sensorRef.dataId)
1806 ccdExposure = sensorRef.get(self.config.datasetType)
1808 camera = sensorRef.get("camera")
1809 isrData = self.readIsrData(sensorRef, ccdExposure)
1811 result = self.run(ccdExposure, camera=camera, **isrData.getDict())
1813 if self.config.doWrite:
1814 sensorRef.put(result.exposure, "postISRCCD")
1815 if result.preInterpExposure is not None:
1816 sensorRef.put(result.preInterpExposure, "postISRCCD_uninterpolated")
1817 if result.ossThumb is not None:
1818 isrQa.writeThumbnail(sensorRef, result.ossThumb, "ossThumb")
1819 if result.flattenedThumb is not None:
1820 isrQa.writeThumbnail(sensorRef, result.flattenedThumb, "flattenedThumb")
1822 return result
1824 def getIsrExposure(self, dataRef, datasetType, dateObs=None, immediate=True):
1825 """Retrieve a calibration dataset for removing instrument signature.
1827 Parameters
1828 ----------
1830 dataRef : `daf.persistence.butlerSubset.ButlerDataRef`
1831 DataRef of the detector data to find calibration datasets
1832 for.
1833 datasetType : `str`
1834 Type of dataset to retrieve (e.g. 'bias', 'flat', etc).
1835 dateObs : `str`, optional
1836 Date of the observation. Used to correct butler failures
1837 when using fallback filters.
1838 immediate : `Bool`
1839 If True, disable butler proxies to enable error handling
1840 within this routine.
1842 Returns
1843 -------
1844 exposure : `lsst.afw.image.Exposure`
1845 Requested calibration frame.
1847 Raises
1848 ------
1849 RuntimeError
1850 Raised if no matching calibration frame can be found.
1851 """
1852 try:
1853 exp = dataRef.get(datasetType, immediate=immediate)
1854 except Exception as exc1:
1855 if not self.config.fallbackFilterName:
1856 raise RuntimeError("Unable to retrieve %s for %s: %s." % (datasetType, dataRef.dataId, exc1))
1857 try:
1858 if self.config.useFallbackDate and dateObs:
1859 exp = dataRef.get(datasetType, filter=self.config.fallbackFilterName,
1860 dateObs=dateObs, immediate=immediate)
1861 else:
1862 exp = dataRef.get(datasetType, filter=self.config.fallbackFilterName, immediate=immediate)
1863 except Exception as exc2:
1864 raise RuntimeError("Unable to retrieve %s for %s, even with fallback filter %s: %s AND %s." %
1865 (datasetType, dataRef.dataId, self.config.fallbackFilterName, exc1, exc2))
1866 self.log.warning("Using fallback calibration from filter %s.", self.config.fallbackFilterName)
1868 if self.config.doAssembleIsrExposures:
1869 exp = self.assembleCcd.assembleCcd(exp)
1870 return exp
1872 def ensureExposure(self, inputExp, camera=None, detectorNum=None):
1873 """Ensure that the data returned by Butler is a fully constructed exp.
1875 ISR requires exposure-level image data for historical reasons, so if we
1876 did not recieve that from Butler, construct it from what we have,
1877 modifying the input in place.
1879 Parameters
1880 ----------
1881 inputExp : `lsst.afw.image.Exposure`, `lsst.afw.image.DecoratedImageU`,
1882 or `lsst.afw.image.ImageF`
1883 The input data structure obtained from Butler.
1884 camera : `lsst.afw.cameraGeom.camera`, optional
1885 The camera associated with the image. Used to find the appropriate
1886 detector if detector is not already set.
1887 detectorNum : `int`, optional
1888 The detector in the camera to attach, if the detector is not
1889 already set.
1891 Returns
1892 -------
1893 inputExp : `lsst.afw.image.Exposure`
1894 The re-constructed exposure, with appropriate detector parameters.
1896 Raises
1897 ------
1898 TypeError
1899 Raised if the input data cannot be used to construct an exposure.
1900 """
1901 if isinstance(inputExp, afwImage.DecoratedImageU):
1902 inputExp = afwImage.makeExposure(afwImage.makeMaskedImage(inputExp))
1903 elif isinstance(inputExp, afwImage.ImageF):
1904 inputExp = afwImage.makeExposure(afwImage.makeMaskedImage(inputExp))
1905 elif isinstance(inputExp, afwImage.MaskedImageF):
1906 inputExp = afwImage.makeExposure(inputExp)
1907 elif isinstance(inputExp, afwImage.Exposure):
1908 pass
1909 elif inputExp is None:
1910 # Assume this will be caught by the setup if it is a problem.
1911 return inputExp
1912 else:
1913 raise TypeError("Input Exposure is not known type in isrTask.ensureExposure: %s." %
1914 (type(inputExp), ))
1916 if inputExp.getDetector() is None:
1917 if camera is None or detectorNum is None:
1918 raise RuntimeError('Must supply both a camera and detector number when using exposures '
1919 'without a detector set.')
1920 inputExp.setDetector(camera[detectorNum])
1922 return inputExp
1924 def convertIntToFloat(self, exposure):
1925 """Convert exposure image from uint16 to float.
1927 If the exposure does not need to be converted, the input is
1928 immediately returned. For exposures that are converted to use
1929 floating point pixels, the variance is set to unity and the
1930 mask to zero.
1932 Parameters
1933 ----------
1934 exposure : `lsst.afw.image.Exposure`
1935 The raw exposure to be converted.
1937 Returns
1938 -------
1939 newexposure : `lsst.afw.image.Exposure`
1940 The input ``exposure``, converted to floating point pixels.
1942 Raises
1943 ------
1944 RuntimeError
1945 Raised if the exposure type cannot be converted to float.
1947 """
1948 if isinstance(exposure, afwImage.ExposureF):
1949 # Nothing to be done
1950 self.log.debug("Exposure already of type float.")
1951 return exposure
1952 if not hasattr(exposure, "convertF"):
1953 raise RuntimeError("Unable to convert exposure (%s) to float." % type(exposure))
1955 newexposure = exposure.convertF()
1956 newexposure.variance[:] = 1
1957 newexposure.mask[:] = 0x0
1959 return newexposure
1961 def maskAmplifier(self, ccdExposure, amp, defects):
1962 """Identify bad amplifiers, saturated and suspect pixels.
1964 Parameters
1965 ----------
1966 ccdExposure : `lsst.afw.image.Exposure`
1967 Input exposure to be masked.
1968 amp : `lsst.afw.table.AmpInfoCatalog`
1969 Catalog of parameters defining the amplifier on this
1970 exposure to mask.
1971 defects : `lsst.ip.isr.Defects`
1972 List of defects. Used to determine if the entire
1973 amplifier is bad.
1975 Returns
1976 -------
1977 badAmp : `Bool`
1978 If this is true, the entire amplifier area is covered by
1979 defects and unusable.
1981 """
1982 maskedImage = ccdExposure.getMaskedImage()
1984 badAmp = False
1986 # Check if entire amp region is defined as a defect
1987 # NB: need to use amp.getBBox() for correct comparison with current
1988 # defects definition.
1989 if defects is not None:
1990 badAmp = bool(sum([v.getBBox().contains(amp.getBBox()) for v in defects]))
1992 # In the case of a bad amp, we will set mask to "BAD"
1993 # (here use amp.getRawBBox() for correct association with pixels in
1994 # current ccdExposure).
1995 if badAmp:
1996 dataView = afwImage.MaskedImageF(maskedImage, amp.getRawBBox(),
1997 afwImage.PARENT)
1998 maskView = dataView.getMask()
1999 maskView |= maskView.getPlaneBitMask("BAD")
2000 del maskView
2001 return badAmp
2003 # Mask remaining defects after assembleCcd() to allow for defects that
2004 # cross amplifier boundaries. Saturation and suspect pixels can be
2005 # masked now, though.
2006 limits = dict()
2007 if self.config.doSaturation and not badAmp:
2008 limits.update({self.config.saturatedMaskName: amp.getSaturation()})
2009 if self.config.doSuspect and not badAmp:
2010 limits.update({self.config.suspectMaskName: amp.getSuspectLevel()})
2011 if math.isfinite(self.config.saturation):
2012 limits.update({self.config.saturatedMaskName: self.config.saturation})
2014 for maskName, maskThreshold in limits.items():
2015 if not math.isnan(maskThreshold):
2016 dataView = maskedImage.Factory(maskedImage, amp.getRawBBox())
2017 isrFunctions.makeThresholdMask(
2018 maskedImage=dataView,
2019 threshold=maskThreshold,
2020 growFootprints=0,
2021 maskName=maskName
2022 )
2024 # Determine if we've fully masked this amplifier with SUSPECT and
2025 # SAT pixels.
2026 maskView = afwImage.Mask(maskedImage.getMask(), amp.getRawDataBBox(),
2027 afwImage.PARENT)
2028 maskVal = maskView.getPlaneBitMask([self.config.saturatedMaskName,
2029 self.config.suspectMaskName])
2030 if numpy.all(maskView.getArray() & maskVal > 0):
2031 badAmp = True
2032 maskView |= maskView.getPlaneBitMask("BAD")
2034 return badAmp
2036 def overscanCorrection(self, ccdExposure, amp):
2037 """Apply overscan correction in place.
2039 This method does initial pixel rejection of the overscan
2040 region. The overscan can also be optionally segmented to
2041 allow for discontinuous overscan responses to be fit
2042 separately. The actual overscan subtraction is performed by
2043 the `lsst.ip.isr.isrFunctions.overscanCorrection` function,
2044 which is called here after the amplifier is preprocessed.
2046 Parameters
2047 ----------
2048 ccdExposure : `lsst.afw.image.Exposure`
2049 Exposure to have overscan correction performed.
2050 amp : `lsst.afw.cameraGeom.Amplifer`
2051 The amplifier to consider while correcting the overscan.
2053 Returns
2054 -------
2055 overscanResults : `lsst.pipe.base.Struct`
2056 Result struct with components:
2057 - ``imageFit`` : scalar or `lsst.afw.image.Image`
2058 Value or fit subtracted from the amplifier image data.
2059 - ``overscanFit`` : scalar or `lsst.afw.image.Image`
2060 Value or fit subtracted from the overscan image data.
2061 - ``overscanImage`` : `lsst.afw.image.Image`
2062 Image of the overscan region with the overscan
2063 correction applied. This quantity is used to estimate
2064 the amplifier read noise empirically.
2066 Raises
2067 ------
2068 RuntimeError
2069 Raised if the ``amp`` does not contain raw pixel information.
2071 See Also
2072 --------
2073 lsst.ip.isr.isrFunctions.overscanCorrection
2074 """
2075 if amp.getRawHorizontalOverscanBBox().isEmpty():
2076 self.log.info("ISR_OSCAN: No overscan region. Not performing overscan correction.")
2077 return None
2079 statControl = afwMath.StatisticsControl()
2080 statControl.setAndMask(ccdExposure.mask.getPlaneBitMask("SAT"))
2082 # Determine the bounding boxes
2083 dataBBox = amp.getRawDataBBox()
2084 oscanBBox = amp.getRawHorizontalOverscanBBox()
2085 dx0 = 0
2086 dx1 = 0
2088 prescanBBox = amp.getRawPrescanBBox()
2089 if (oscanBBox.getBeginX() > prescanBBox.getBeginX()): # amp is at the right
2090 dx0 += self.config.overscanNumLeadingColumnsToSkip
2091 dx1 -= self.config.overscanNumTrailingColumnsToSkip
2092 else:
2093 dx0 += self.config.overscanNumTrailingColumnsToSkip
2094 dx1 -= self.config.overscanNumLeadingColumnsToSkip
2096 # Determine if we need to work on subregions of the amplifier
2097 # and overscan.
2098 imageBBoxes = []
2099 overscanBBoxes = []
2101 if ((self.config.overscanBiasJump
2102 and self.config.overscanBiasJumpLocation)
2103 and (ccdExposure.getMetadata().exists(self.config.overscanBiasJumpKeyword)
2104 and ccdExposure.getMetadata().getScalar(self.config.overscanBiasJumpKeyword) in
2105 self.config.overscanBiasJumpDevices)):
2106 if amp.getReadoutCorner() in (ReadoutCorner.LL, ReadoutCorner.LR):
2107 yLower = self.config.overscanBiasJumpLocation
2108 yUpper = dataBBox.getHeight() - yLower
2109 else:
2110 yUpper = self.config.overscanBiasJumpLocation
2111 yLower = dataBBox.getHeight() - yUpper
2113 imageBBoxes.append(lsst.geom.Box2I(dataBBox.getBegin(),
2114 lsst.geom.Extent2I(dataBBox.getWidth(), yLower)))
2115 overscanBBoxes.append(lsst.geom.Box2I(oscanBBox.getBegin() + lsst.geom.Extent2I(dx0, 0),
2116 lsst.geom.Extent2I(oscanBBox.getWidth() - dx0 + dx1,
2117 yLower)))
2119 imageBBoxes.append(lsst.geom.Box2I(dataBBox.getBegin() + lsst.geom.Extent2I(0, yLower),
2120 lsst.geom.Extent2I(dataBBox.getWidth(), yUpper)))
2121 overscanBBoxes.append(lsst.geom.Box2I(oscanBBox.getBegin() + lsst.geom.Extent2I(dx0, yLower),
2122 lsst.geom.Extent2I(oscanBBox.getWidth() - dx0 + dx1,
2123 yUpper)))
2124 else:
2125 imageBBoxes.append(lsst.geom.Box2I(dataBBox.getBegin(),
2126 lsst.geom.Extent2I(dataBBox.getWidth(), dataBBox.getHeight())))
2127 overscanBBoxes.append(lsst.geom.Box2I(oscanBBox.getBegin() + lsst.geom.Extent2I(dx0, 0),
2128 lsst.geom.Extent2I(oscanBBox.getWidth() - dx0 + dx1,
2129 oscanBBox.getHeight())))
2131 # Perform overscan correction on subregions, ensuring saturated
2132 # pixels are masked.
2133 for imageBBox, overscanBBox in zip(imageBBoxes, overscanBBoxes):
2134 ampImage = ccdExposure.maskedImage[imageBBox]
2135 overscanImage = ccdExposure.maskedImage[overscanBBox]
2137 overscanArray = overscanImage.image.array
2138 median = numpy.ma.median(numpy.ma.masked_where(overscanImage.mask.array, overscanArray))
2139 bad = numpy.where(numpy.abs(overscanArray - median) > self.config.overscanMaxDev)
2140 overscanImage.mask.array[bad] = overscanImage.mask.getPlaneBitMask("SAT")
2142 statControl = afwMath.StatisticsControl()
2143 statControl.setAndMask(ccdExposure.mask.getPlaneBitMask("SAT"))
2145 overscanResults = self.overscan.run(ampImage.getImage(), overscanImage, amp)
2147 # Measure average overscan levels and record them in the metadata.
2148 levelStat = afwMath.MEDIAN
2149 sigmaStat = afwMath.STDEVCLIP
2151 sctrl = afwMath.StatisticsControl(self.config.qa.flatness.clipSigma,
2152 self.config.qa.flatness.nIter)
2153 metadata = ccdExposure.getMetadata()
2154 ampNum = amp.getName()
2155 # if self.config.overscanFitType in ("MEDIAN", "MEAN", "MEANCLIP"):
2156 if isinstance(overscanResults.overscanFit, float):
2157 metadata.set("ISR_OSCAN_LEVEL%s" % ampNum, overscanResults.overscanFit)
2158 metadata.set("ISR_OSCAN_SIGMA%s" % ampNum, 0.0)
2159 else:
2160 stats = afwMath.makeStatistics(overscanResults.overscanFit, levelStat | sigmaStat, sctrl)
2161 metadata.set("ISR_OSCAN_LEVEL%s" % ampNum, stats.getValue(levelStat))
2162 metadata.set("ISR_OSCAN_SIGMA%s" % ampNum, stats.getValue(sigmaStat))
2164 return overscanResults
2166 def updateVariance(self, ampExposure, amp, overscanImage=None, ptcDataset=None):
2167 """Set the variance plane using the gain and read noise
2169 The read noise is calculated from the ``overscanImage`` if the
2170 ``doEmpiricalReadNoise`` option is set in the configuration; otherwise
2171 the value from the amplifier data is used.
2173 Parameters
2174 ----------
2175 ampExposure : `lsst.afw.image.Exposure`
2176 Exposure to process.
2177 amp : `lsst.afw.table.AmpInfoRecord` or `FakeAmp`
2178 Amplifier detector data.
2179 overscanImage : `lsst.afw.image.MaskedImage`, optional.
2180 Image of overscan, required only for empirical read noise.
2181 ptcDataset : `lsst.ip.isr.PhotonTransferCurveDataset`, optional
2182 PTC dataset containing the gains and read noise.
2185 Raises
2186 ------
2187 RuntimeError
2188 Raised if either ``usePtcGains`` of ``usePtcReadNoise``
2189 are ``True``, but ptcDataset is not provided.
2191 Raised if ```doEmpiricalReadNoise`` is ``True`` but
2192 ``overscanImage`` is ``None``.
2194 See also
2195 --------
2196 lsst.ip.isr.isrFunctions.updateVariance
2197 """
2198 maskPlanes = [self.config.saturatedMaskName, self.config.suspectMaskName]
2199 if self.config.usePtcGains:
2200 if ptcDataset is None:
2201 raise RuntimeError("No ptcDataset provided to use PTC gains.")
2202 else:
2203 gain = ptcDataset.gain[amp.getName()]
2204 self.log.info("Using gain from Photon Transfer Curve.")
2205 else:
2206 gain = amp.getGain()
2208 if math.isnan(gain):
2209 gain = 1.0
2210 self.log.warning("Gain set to NAN! Updating to 1.0 to generate Poisson variance.")
2211 elif gain <= 0:
2212 patchedGain = 1.0
2213 self.log.warning("Gain for amp %s == %g <= 0; setting to %f.",
2214 amp.getName(), gain, patchedGain)
2215 gain = patchedGain
2217 if self.config.doEmpiricalReadNoise and overscanImage is None:
2218 raise RuntimeError("Overscan is none for EmpiricalReadNoise.")
2220 if self.config.doEmpiricalReadNoise and overscanImage is not None:
2221 stats = afwMath.StatisticsControl()
2222 stats.setAndMask(overscanImage.mask.getPlaneBitMask(maskPlanes))
2223 readNoise = afwMath.makeStatistics(overscanImage, afwMath.STDEVCLIP, stats).getValue()
2224 self.log.info("Calculated empirical read noise for amp %s: %f.",
2225 amp.getName(), readNoise)
2226 elif self.config.usePtcReadNoise:
2227 if ptcDataset is None:
2228 raise RuntimeError("No ptcDataset provided to use PTC readnoise.")
2229 else:
2230 readNoise = ptcDataset.noise[amp.getName()]
2231 self.log.info("Using read noise from Photon Transfer Curve.")
2232 else:
2233 readNoise = amp.getReadNoise()
2235 isrFunctions.updateVariance(
2236 maskedImage=ampExposure.getMaskedImage(),
2237 gain=gain,
2238 readNoise=readNoise,
2239 )
2241 def maskNegativeVariance(self, exposure):
2242 """Identify and mask pixels with negative variance values.
2244 Parameters
2245 ----------
2246 exposure : `lsst.afw.image.Exposure`
2247 Exposure to process.
2249 See Also
2250 --------
2251 lsst.ip.isr.isrFunctions.updateVariance
2252 """
2253 maskPlane = exposure.getMask().getPlaneBitMask(self.config.negativeVarianceMaskName)
2254 bad = numpy.where(exposure.getVariance().getArray() <= 0.0)
2255 exposure.mask.array[bad] |= maskPlane
2257 def darkCorrection(self, exposure, darkExposure, invert=False):
2258 """Apply dark correction in place.
2260 Parameters
2261 ----------
2262 exposure : `lsst.afw.image.Exposure`
2263 Exposure to process.
2264 darkExposure : `lsst.afw.image.Exposure`
2265 Dark exposure of the same size as ``exposure``.
2266 invert : `Bool`, optional
2267 If True, re-add the dark to an already corrected image.
2269 Raises
2270 ------
2271 RuntimeError
2272 Raised if either ``exposure`` or ``darkExposure`` do not
2273 have their dark time defined.
2275 See Also
2276 --------
2277 lsst.ip.isr.isrFunctions.darkCorrection
2278 """
2279 expScale = exposure.getInfo().getVisitInfo().getDarkTime()
2280 if math.isnan(expScale):
2281 raise RuntimeError("Exposure darktime is NAN.")
2282 if darkExposure.getInfo().getVisitInfo() is not None \
2283 and not math.isnan(darkExposure.getInfo().getVisitInfo().getDarkTime()):
2284 darkScale = darkExposure.getInfo().getVisitInfo().getDarkTime()
2285 else:
2286 # DM-17444: darkExposure.getInfo.getVisitInfo() is None
2287 # so getDarkTime() does not exist.
2288 self.log.warning("darkExposure.getInfo().getVisitInfo() does not exist. Using darkScale = 1.0.")
2289 darkScale = 1.0
2291 isrFunctions.darkCorrection(
2292 maskedImage=exposure.getMaskedImage(),
2293 darkMaskedImage=darkExposure.getMaskedImage(),
2294 expScale=expScale,
2295 darkScale=darkScale,
2296 invert=invert,
2297 trimToFit=self.config.doTrimToMatchCalib
2298 )
2300 def doLinearize(self, detector):
2301 """Check if linearization is needed for the detector cameraGeom.
2303 Checks config.doLinearize and the linearity type of the first
2304 amplifier.
2306 Parameters
2307 ----------
2308 detector : `lsst.afw.cameraGeom.Detector`
2309 Detector to get linearity type from.
2311 Returns
2312 -------
2313 doLinearize : `Bool`
2314 If True, linearization should be performed.
2315 """
2316 return self.config.doLinearize and \
2317 detector.getAmplifiers()[0].getLinearityType() != NullLinearityType
2319 def flatCorrection(self, exposure, flatExposure, invert=False):
2320 """Apply flat correction in place.
2322 Parameters
2323 ----------
2324 exposure : `lsst.afw.image.Exposure`
2325 Exposure to process.
2326 flatExposure : `lsst.afw.image.Exposure`
2327 Flat exposure of the same size as ``exposure``.
2328 invert : `Bool`, optional
2329 If True, unflatten an already flattened image.
2331 See Also
2332 --------
2333 lsst.ip.isr.isrFunctions.flatCorrection
2334 """
2335 isrFunctions.flatCorrection(
2336 maskedImage=exposure.getMaskedImage(),
2337 flatMaskedImage=flatExposure.getMaskedImage(),
2338 scalingType=self.config.flatScalingType,
2339 userScale=self.config.flatUserScale,
2340 invert=invert,
2341 trimToFit=self.config.doTrimToMatchCalib
2342 )
2344 def saturationDetection(self, exposure, amp):
2345 """Detect and mask saturated pixels in config.saturatedMaskName.
2347 Parameters
2348 ----------
2349 exposure : `lsst.afw.image.Exposure`
2350 Exposure to process. Only the amplifier DataSec is processed.
2351 amp : `lsst.afw.table.AmpInfoCatalog`
2352 Amplifier detector data.
2354 See Also
2355 --------
2356 lsst.ip.isr.isrFunctions.makeThresholdMask
2357 """
2358 if not math.isnan(amp.getSaturation()):
2359 maskedImage = exposure.getMaskedImage()
2360 dataView = maskedImage.Factory(maskedImage, amp.getRawBBox())
2361 isrFunctions.makeThresholdMask(
2362 maskedImage=dataView,
2363 threshold=amp.getSaturation(),
2364 growFootprints=0,
2365 maskName=self.config.saturatedMaskName,
2366 )
2368 def saturationInterpolation(self, exposure):
2369 """Interpolate over saturated pixels, in place.
2371 This method should be called after `saturationDetection`, to
2372 ensure that the saturated pixels have been identified in the
2373 SAT mask. It should also be called after `assembleCcd`, since
2374 saturated regions may cross amplifier boundaries.
2376 Parameters
2377 ----------
2378 exposure : `lsst.afw.image.Exposure`
2379 Exposure to process.
2381 See Also
2382 --------
2383 lsst.ip.isr.isrTask.saturationDetection
2384 lsst.ip.isr.isrFunctions.interpolateFromMask
2385 """
2386 isrFunctions.interpolateFromMask(
2387 maskedImage=exposure.getMaskedImage(),
2388 fwhm=self.config.fwhm,
2389 growSaturatedFootprints=self.config.growSaturationFootprintSize,
2390 maskNameList=list(self.config.saturatedMaskName),
2391 )
2393 def suspectDetection(self, exposure, amp):
2394 """Detect and mask suspect pixels in config.suspectMaskName.
2396 Parameters
2397 ----------
2398 exposure : `lsst.afw.image.Exposure`
2399 Exposure to process. Only the amplifier DataSec is processed.
2400 amp : `lsst.afw.table.AmpInfoCatalog`
2401 Amplifier detector data.
2403 See Also
2404 --------
2405 lsst.ip.isr.isrFunctions.makeThresholdMask
2407 Notes
2408 -----
2409 Suspect pixels are pixels whose value is greater than
2410 amp.getSuspectLevel(). This is intended to indicate pixels that may be
2411 affected by unknown systematics; for example if non-linearity
2412 corrections above a certain level are unstable then that would be a
2413 useful value for suspectLevel. A value of `nan` indicates that no such
2414 level exists and no pixels are to be masked as suspicious.
2415 """
2416 suspectLevel = amp.getSuspectLevel()
2417 if math.isnan(suspectLevel):
2418 return
2420 maskedImage = exposure.getMaskedImage()
2421 dataView = maskedImage.Factory(maskedImage, amp.getRawBBox())
2422 isrFunctions.makeThresholdMask(
2423 maskedImage=dataView,
2424 threshold=suspectLevel,
2425 growFootprints=0,
2426 maskName=self.config.suspectMaskName,
2427 )
2429 def maskDefect(self, exposure, defectBaseList):
2430 """Mask defects using mask plane "BAD", in place.
2432 Parameters
2433 ----------
2434 exposure : `lsst.afw.image.Exposure`
2435 Exposure to process.
2436 defectBaseList : `lsst.ip.isr.Defects` or `list` of
2437 `lsst.afw.image.DefectBase`.
2438 List of defects to mask.
2440 Notes
2441 -----
2442 Call this after CCD assembly, since defects may cross amplifier
2443 boundaries.
2444 """
2445 maskedImage = exposure.getMaskedImage()
2446 if not isinstance(defectBaseList, Defects):
2447 # Promotes DefectBase to Defect
2448 defectList = Defects(defectBaseList)
2449 else:
2450 defectList = defectBaseList
2451 defectList.maskPixels(maskedImage, maskName="BAD")
2453 def maskEdges(self, exposure, numEdgePixels=0, maskPlane="SUSPECT", level='DETECTOR'):
2454 """Mask edge pixels with applicable mask plane.
2456 Parameters
2457 ----------
2458 exposure : `lsst.afw.image.Exposure`
2459 Exposure to process.
2460 numEdgePixels : `int`, optional
2461 Number of edge pixels to mask.
2462 maskPlane : `str`, optional
2463 Mask plane name to use.
2464 level : `str`, optional
2465 Level at which to mask edges.
2466 """
2467 maskedImage = exposure.getMaskedImage()
2468 maskBitMask = maskedImage.getMask().getPlaneBitMask(maskPlane)
2470 if numEdgePixels > 0:
2471 if level == 'DETECTOR':
2472 boxes = [maskedImage.getBBox()]
2473 elif level == 'AMP':
2474 boxes = [amp.getBBox() for amp in exposure.getDetector()]
2476 for box in boxes:
2477 # This makes a bbox numEdgeSuspect pixels smaller than the
2478 # image on each side
2479 subImage = maskedImage[box]
2480 box.grow(-numEdgePixels)
2481 # Mask pixels outside box
2482 SourceDetectionTask.setEdgeBits(
2483 subImage,
2484 box,
2485 maskBitMask)
2487 def maskAndInterpolateDefects(self, exposure, defectBaseList):
2488 """Mask and interpolate defects using mask plane "BAD", in place.
2490 Parameters
2491 ----------
2492 exposure : `lsst.afw.image.Exposure`
2493 Exposure to process.
2494 defectBaseList : `lsst.ip.isr.Defects` or `list` of
2495 `lsst.afw.image.DefectBase`.
2496 List of defects to mask and interpolate.
2498 See Also
2499 --------
2500 lsst.ip.isr.isrTask.maskDefect
2501 """
2502 self.maskDefect(exposure, defectBaseList)
2503 self.maskEdges(exposure, numEdgePixels=self.config.numEdgeSuspect,
2504 maskPlane="SUSPECT", level=self.config.edgeMaskLevel)
2505 isrFunctions.interpolateFromMask(
2506 maskedImage=exposure.getMaskedImage(),
2507 fwhm=self.config.fwhm,
2508 growSaturatedFootprints=0,
2509 maskNameList=["BAD"],
2510 )
2512 def maskNan(self, exposure):
2513 """Mask NaNs using mask plane "UNMASKEDNAN", in place.
2515 Parameters
2516 ----------
2517 exposure : `lsst.afw.image.Exposure`
2518 Exposure to process.
2520 Notes
2521 -----
2522 We mask over all non-finite values (NaN, inf), including those
2523 that are masked with other bits (because those may or may not be
2524 interpolated over later, and we want to remove all NaN/infs).
2525 Despite this behaviour, the "UNMASKEDNAN" mask plane is used to
2526 preserve the historical name.
2527 """
2528 maskedImage = exposure.getMaskedImage()
2530 # Find and mask NaNs
2531 maskedImage.getMask().addMaskPlane("UNMASKEDNAN")
2532 maskVal = maskedImage.getMask().getPlaneBitMask("UNMASKEDNAN")
2533 numNans = maskNans(maskedImage, maskVal)
2534 self.metadata.set("NUMNANS", numNans)
2535 if numNans > 0:
2536 self.log.warning("There were %d unmasked NaNs.", numNans)
2538 def maskAndInterpolateNan(self, exposure):
2539 """"Mask and interpolate NaN/infs using mask plane "UNMASKEDNAN",
2540 in place.
2542 Parameters
2543 ----------
2544 exposure : `lsst.afw.image.Exposure`
2545 Exposure to process.
2547 See Also
2548 --------
2549 lsst.ip.isr.isrTask.maskNan
2550 """
2551 self.maskNan(exposure)
2552 isrFunctions.interpolateFromMask(
2553 maskedImage=exposure.getMaskedImage(),
2554 fwhm=self.config.fwhm,
2555 growSaturatedFootprints=0,
2556 maskNameList=["UNMASKEDNAN"],
2557 )
2559 def measureBackground(self, exposure, IsrQaConfig=None):
2560 """Measure the image background in subgrids, for quality control.
2562 Parameters
2563 ----------
2564 exposure : `lsst.afw.image.Exposure`
2565 Exposure to process.
2566 IsrQaConfig : `lsst.ip.isr.isrQa.IsrQaConfig`
2567 Configuration object containing parameters on which background
2568 statistics and subgrids to use.
2569 """
2570 if IsrQaConfig is not None:
2571 statsControl = afwMath.StatisticsControl(IsrQaConfig.flatness.clipSigma,
2572 IsrQaConfig.flatness.nIter)
2573 maskVal = exposure.getMaskedImage().getMask().getPlaneBitMask(["BAD", "SAT", "DETECTED"])
2574 statsControl.setAndMask(maskVal)
2575 maskedImage = exposure.getMaskedImage()
2576 stats = afwMath.makeStatistics(maskedImage, afwMath.MEDIAN | afwMath.STDEVCLIP, statsControl)
2577 skyLevel = stats.getValue(afwMath.MEDIAN)
2578 skySigma = stats.getValue(afwMath.STDEVCLIP)
2579 self.log.info("Flattened sky level: %f +/- %f.", skyLevel, skySigma)
2580 metadata = exposure.getMetadata()
2581 metadata.set('SKYLEVEL', skyLevel)
2582 metadata.set('SKYSIGMA', skySigma)
2584 # calcluating flatlevel over the subgrids
2585 stat = afwMath.MEANCLIP if IsrQaConfig.flatness.doClip else afwMath.MEAN
2586 meshXHalf = int(IsrQaConfig.flatness.meshX/2.)
2587 meshYHalf = int(IsrQaConfig.flatness.meshY/2.)
2588 nX = int((exposure.getWidth() + meshXHalf) / IsrQaConfig.flatness.meshX)
2589 nY = int((exposure.getHeight() + meshYHalf) / IsrQaConfig.flatness.meshY)
2590 skyLevels = numpy.zeros((nX, nY))
2592 for j in range(nY):
2593 yc = meshYHalf + j * IsrQaConfig.flatness.meshY
2594 for i in range(nX):
2595 xc = meshXHalf + i * IsrQaConfig.flatness.meshX
2597 xLLC = xc - meshXHalf
2598 yLLC = yc - meshYHalf
2599 xURC = xc + meshXHalf - 1
2600 yURC = yc + meshYHalf - 1
2602 bbox = lsst.geom.Box2I(lsst.geom.Point2I(xLLC, yLLC), lsst.geom.Point2I(xURC, yURC))
2603 miMesh = maskedImage.Factory(exposure.getMaskedImage(), bbox, afwImage.LOCAL)
2605 skyLevels[i, j] = afwMath.makeStatistics(miMesh, stat, statsControl).getValue()
2607 good = numpy.where(numpy.isfinite(skyLevels))
2608 skyMedian = numpy.median(skyLevels[good])
2609 flatness = (skyLevels[good] - skyMedian) / skyMedian
2610 flatness_rms = numpy.std(flatness)
2611 flatness_pp = flatness.max() - flatness.min() if len(flatness) > 0 else numpy.nan
2613 self.log.info("Measuring sky levels in %dx%d grids: %f.", nX, nY, skyMedian)
2614 self.log.info("Sky flatness in %dx%d grids - pp: %f rms: %f.",
2615 nX, nY, flatness_pp, flatness_rms)
2617 metadata.set('FLATNESS_PP', float(flatness_pp))
2618 metadata.set('FLATNESS_RMS', float(flatness_rms))
2619 metadata.set('FLATNESS_NGRIDS', '%dx%d' % (nX, nY))
2620 metadata.set('FLATNESS_MESHX', IsrQaConfig.flatness.meshX)
2621 metadata.set('FLATNESS_MESHY', IsrQaConfig.flatness.meshY)
2623 def roughZeroPoint(self, exposure):
2624 """Set an approximate magnitude zero point for the exposure.
2626 Parameters
2627 ----------
2628 exposure : `lsst.afw.image.Exposure`
2629 Exposure to process.
2630 """
2631 filterLabel = exposure.getFilterLabel()
2632 physicalFilter = isrFunctions.getPhysicalFilter(filterLabel, self.log)
2634 if physicalFilter in self.config.fluxMag0T1:
2635 fluxMag0 = self.config.fluxMag0T1[physicalFilter]
2636 else:
2637 self.log.warning("No rough magnitude zero point defined for filter %s.", physicalFilter)
2638 fluxMag0 = self.config.defaultFluxMag0T1
2640 expTime = exposure.getInfo().getVisitInfo().getExposureTime()
2641 if not expTime > 0: # handle NaN as well as <= 0
2642 self.log.warning("Non-positive exposure time; skipping rough zero point.")
2643 return
2645 self.log.info("Setting rough magnitude zero point for filter %s: %f",
2646 physicalFilter, 2.5*math.log10(fluxMag0*expTime))
2647 exposure.setPhotoCalib(afwImage.makePhotoCalibFromCalibZeroPoint(fluxMag0*expTime, 0.0))
2649 def setValidPolygonIntersect(self, ccdExposure, fpPolygon):
2650 """Set valid polygon as the intersection of fpPolygon and chip corners.
2652 Parameters
2653 ----------
2654 ccdExposure : `lsst.afw.image.Exposure`
2655 Exposure to process.
2656 fpPolygon : `lsst.afw.geom.Polygon`
2657 Polygon in focal plane coordinates.
2658 """
2659 # Get ccd corners in focal plane coordinates
2660 ccd = ccdExposure.getDetector()
2661 fpCorners = ccd.getCorners(FOCAL_PLANE)
2662 ccdPolygon = Polygon(fpCorners)
2664 # Get intersection of ccd corners with fpPolygon
2665 intersect = ccdPolygon.intersectionSingle(fpPolygon)
2667 # Transform back to pixel positions and build new polygon
2668 ccdPoints = ccd.transform(intersect, FOCAL_PLANE, PIXELS)
2669 validPolygon = Polygon(ccdPoints)
2670 ccdExposure.getInfo().setValidPolygon(validPolygon)
2672 @contextmanager
2673 def flatContext(self, exp, flat, dark=None):
2674 """Context manager that applies and removes flats and darks,
2675 if the task is configured to apply them.
2677 Parameters
2678 ----------
2679 exp : `lsst.afw.image.Exposure`
2680 Exposure to process.
2681 flat : `lsst.afw.image.Exposure`
2682 Flat exposure the same size as ``exp``.
2683 dark : `lsst.afw.image.Exposure`, optional
2684 Dark exposure the same size as ``exp``.
2686 Yields
2687 ------
2688 exp : `lsst.afw.image.Exposure`
2689 The flat and dark corrected exposure.
2690 """
2691 if self.config.doDark and dark is not None:
2692 self.darkCorrection(exp, dark)
2693 if self.config.doFlat:
2694 self.flatCorrection(exp, flat)
2695 try:
2696 yield exp
2697 finally:
2698 if self.config.doFlat:
2699 self.flatCorrection(exp, flat, invert=True)
2700 if self.config.doDark and dark is not None:
2701 self.darkCorrection(exp, dark, invert=True)
2703 def debugView(self, exposure, stepname):
2704 """Utility function to examine ISR exposure at different stages.
2706 Parameters
2707 ----------
2708 exposure : `lsst.afw.image.Exposure`
2709 Exposure to view.
2710 stepname : `str`
2711 State of processing to view.
2712 """
2713 frame = getDebugFrame(self._display, stepname)
2714 if frame:
2715 display = getDisplay(frame)
2716 display.scale('asinh', 'zscale')
2717 display.mtv(exposure)
2718 prompt = "Press Enter to continue [c]... "
2719 while True:
2720 ans = input(prompt).lower()
2721 if ans in ("", "c",):
2722 break
2725class FakeAmp(object):
2726 """A Detector-like object that supports returning gain and saturation level
2728 This is used when the input exposure does not have a detector.
2730 Parameters
2731 ----------
2732 exposure : `lsst.afw.image.Exposure`
2733 Exposure to generate a fake amplifier for.
2734 config : `lsst.ip.isr.isrTaskConfig`
2735 Configuration to apply to the fake amplifier.
2736 """
2738 def __init__(self, exposure, config):
2739 self._bbox = exposure.getBBox(afwImage.LOCAL)
2740 self._RawHorizontalOverscanBBox = lsst.geom.Box2I()
2741 self._gain = config.gain
2742 self._readNoise = config.readNoise
2743 self._saturation = config.saturation
2745 def getBBox(self):
2746 return self._bbox
2748 def getRawBBox(self):
2749 return self._bbox
2751 def getRawHorizontalOverscanBBox(self):
2752 return self._RawHorizontalOverscanBBox
2754 def getGain(self):
2755 return self._gain
2757 def getReadNoise(self):
2758 return self._readNoise
2760 def getSaturation(self):
2761 return self._saturation
2763 def getSuspectLevel(self):
2764 return float("NaN")
2767class RunIsrConfig(pexConfig.Config):
2768 isr = pexConfig.ConfigurableField(target=IsrTask, doc="Instrument signature removal")
2771class RunIsrTask(pipeBase.CmdLineTask):
2772 """Task to wrap the default IsrTask to allow it to be retargeted.
2774 The standard IsrTask can be called directly from a command line
2775 program, but doing so removes the ability of the task to be
2776 retargeted. As most cameras override some set of the IsrTask
2777 methods, this would remove those data-specific methods in the
2778 output post-ISR images. This wrapping class fixes the issue,
2779 allowing identical post-ISR images to be generated by both the
2780 processCcd and isrTask code.
2781 """
2782 ConfigClass = RunIsrConfig
2783 _DefaultName = "runIsr"
2785 def __init__(self, *args, **kwargs):
2786 super().__init__(*args, **kwargs)
2787 self.makeSubtask("isr")
2789 def runDataRef(self, dataRef):
2790 """
2791 Parameters
2792 ----------
2793 dataRef : `lsst.daf.persistence.ButlerDataRef`
2794 data reference of the detector data to be processed
2796 Returns
2797 -------
2798 result : `pipeBase.Struct`
2799 Result struct with component:
2801 - exposure : `lsst.afw.image.Exposure`
2802 Post-ISR processed exposure.
2803 """
2804 return self.isr.runDataRef(dataRef)