Hide keyboard shortcuts

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/>. 

21 

22import math 

23import numpy 

24 

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 

31 

32from contextlib import contextmanager 

33from lsstDebug import getDebugFrame 

34 

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 

42 

43from . import isrFunctions 

44from . import isrQa 

45from . import linearize 

46from .defects import Defects 

47 

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 lsst.daf.butler import DimensionGraph 

57 

58 

59__all__ = ["IsrTask", "IsrTaskConfig", "RunIsrTask", "RunIsrConfig"] 

60 

61 

62def crosstalkSourceLookup(datasetType, registry, quantumDataId, collections): 

63 """Lookup function to identify crosstalkSource entries. 

64 

65 This should return an empty list under most circumstances. Only 

66 when inter-chip crosstalk has been identified should this be 

67 populated. 

68 

69 This will be unused until DM-25348 resolves the quantum graph 

70 generation issue. 

71 

72 Parameters 

73 ---------- 

74 datasetType : `str` 

75 Dataset to lookup. 

76 registry : `lsst.daf.butler.Registry` 

77 Butler registry to query. 

78 quantumDataId : `lsst.daf.butler.ExpandedDataCoordinate` 

79 Data id to transform to identify crosstalkSources. The 

80 ``detector`` entry will be stripped. 

81 collections : `lsst.daf.butler.CollectionSearch` 

82 Collections to search through. 

83 

84 Returns 

85 ------- 

86 results : `list` [`lsst.daf.butler.DatasetRef`] 

87 List of datasets that match the query that will be used as 

88 crosstalkSources. 

89 """ 

90 newDataId = quantumDataId.subset(DimensionGraph(registry.dimensions, names=["instrument", "exposure"])) 

91 results = list(registry.queryDatasets(datasetType, 

92 collections=collections, 

93 dataId=newDataId, 

94 findFirst=True, 

95 ).expanded()) 

96 return results 

97 

98 

99class IsrTaskConnections(pipeBase.PipelineTaskConnections, 

100 dimensions={"instrument", "exposure", "detector"}, 

101 defaultTemplates={}): 

102 ccdExposure = cT.Input( 

103 name="raw", 

104 doc="Input exposure to process.", 

105 storageClass="Exposure", 

106 dimensions=["instrument", "exposure", "detector"], 

107 ) 

108 camera = cT.PrerequisiteInput( 

109 name="camera", 

110 storageClass="Camera", 

111 doc="Input camera to construct complete exposures.", 

112 dimensions=["instrument"], 

113 isCalibration=True, 

114 ) 

115 

116 crosstalk = cT.PrerequisiteInput( 

117 name="crosstalk", 

118 doc="Input crosstalk object", 

119 storageClass="CrosstalkCalib", 

120 dimensions=["instrument", "detector"], 

121 isCalibration=True, 

122 ) 

123 # TODO: DM-25348. This does not work yet to correctly load 

124 # possible crosstalk sources. 

125 crosstalkSources = cT.PrerequisiteInput( 

126 name="isrOverscanCorrected", 

127 doc="Overscan corrected input images.", 

128 storageClass="Exposure", 

129 dimensions=["instrument", "exposure", "detector"], 

130 deferLoad=True, 

131 multiple=True, 

132 lookupFunction=crosstalkSourceLookup, 

133 ) 

134 bias = cT.PrerequisiteInput( 

135 name="bias", 

136 doc="Input bias calibration.", 

137 storageClass="ExposureF", 

138 dimensions=["instrument", "detector"], 

139 isCalibration=True, 

140 ) 

141 dark = cT.PrerequisiteInput( 

142 name='dark', 

143 doc="Input dark calibration.", 

144 storageClass="ExposureF", 

145 dimensions=["instrument", "detector"], 

146 isCalibration=True, 

147 ) 

148 flat = cT.PrerequisiteInput( 

149 name="flat", 

150 doc="Input flat calibration.", 

151 storageClass="ExposureF", 

152 dimensions=["instrument", "physical_filter", "detector"], 

153 isCalibration=True, 

154 ) 

155 fringes = cT.PrerequisiteInput( 

156 name="fringe", 

157 doc="Input fringe calibration.", 

158 storageClass="ExposureF", 

159 dimensions=["instrument", "physical_filter", "detector"], 

160 isCalibration=True, 

161 ) 

162 strayLightData = cT.PrerequisiteInput( 

163 name='yBackground', 

164 doc="Input stray light calibration.", 

165 storageClass="StrayLightData", 

166 dimensions=["instrument", "physical_filter", "detector"], 

167 isCalibration=True, 

168 ) 

169 bfKernel = cT.PrerequisiteInput( 

170 name='bfKernel', 

171 doc="Input brighter-fatter kernel.", 

172 storageClass="NumpyArray", 

173 dimensions=["instrument"], 

174 isCalibration=True, 

175 ) 

176 newBFKernel = cT.PrerequisiteInput( 

177 name='brighterFatterKernel', 

178 doc="Newer complete kernel + gain solutions.", 

179 storageClass="BrighterFatterKernel", 

180 dimensions=["instrument", "detector"], 

181 isCalibration=True, 

182 ) 

183 defects = cT.PrerequisiteInput( 

184 name='defects', 

185 doc="Input defect tables.", 

186 storageClass="Defects", 

187 dimensions=["instrument", "detector"], 

188 isCalibration=True, 

189 ) 

190 opticsTransmission = cT.PrerequisiteInput( 

191 name="transmission_optics", 

192 storageClass="TransmissionCurve", 

193 doc="Transmission curve due to the optics.", 

194 dimensions=["instrument"], 

195 isCalibration=True, 

196 ) 

197 filterTransmission = cT.PrerequisiteInput( 

198 name="transmission_filter", 

199 storageClass="TransmissionCurve", 

200 doc="Transmission curve due to the filter.", 

201 dimensions=["instrument", "physical_filter"], 

202 isCalibration=True, 

203 ) 

204 sensorTransmission = cT.PrerequisiteInput( 

205 name="transmission_sensor", 

206 storageClass="TransmissionCurve", 

207 doc="Transmission curve due to the sensor.", 

208 dimensions=["instrument", "detector"], 

209 isCalibration=True, 

210 ) 

211 atmosphereTransmission = cT.PrerequisiteInput( 

212 name="transmission_atmosphere", 

213 storageClass="TransmissionCurve", 

214 doc="Transmission curve due to the atmosphere.", 

215 dimensions=["instrument"], 

216 isCalibration=True, 

217 ) 

218 illumMaskedImage = cT.PrerequisiteInput( 

219 name="illum", 

220 doc="Input illumination correction.", 

221 storageClass="MaskedImageF", 

222 dimensions=["instrument", "physical_filter", "detector"], 

223 isCalibration=True, 

224 ) 

225 

226 outputExposure = cT.Output( 

227 name='postISRCCD', 

228 doc="Output ISR processed exposure.", 

229 storageClass="Exposure", 

230 dimensions=["instrument", "exposure", "detector"], 

231 ) 

232 preInterpExposure = cT.Output( 

233 name='preInterpISRCCD', 

234 doc="Output ISR processed exposure, with pixels left uninterpolated.", 

235 storageClass="ExposureF", 

236 dimensions=["instrument", "exposure", "detector"], 

237 ) 

238 outputOssThumbnail = cT.Output( 

239 name="OssThumb", 

240 doc="Output Overscan-subtracted thumbnail image.", 

241 storageClass="Thumbnail", 

242 dimensions=["instrument", "exposure", "detector"], 

243 ) 

244 outputFlattenedThumbnail = cT.Output( 

245 name="FlattenedThumb", 

246 doc="Output flat-corrected thumbnail image.", 

247 storageClass="Thumbnail", 

248 dimensions=["instrument", "exposure", "detector"], 

249 ) 

250 

251 def __init__(self, *, config=None): 

252 super().__init__(config=config) 

253 

254 if config.doBias is not True: 

255 self.prerequisiteInputs.discard("bias") 

256 if config.doLinearize is not True: 

257 self.prerequisiteInputs.discard("linearizer") 

258 if config.doCrosstalk is not True: 

259 self.inputs.discard("crosstalkSources") 

260 self.prerequisiteInputs.discard("crosstalk") 

261 if config.doBrighterFatter is not True: 

262 self.prerequisiteInputs.discard("bfKernel") 

263 self.prerequisiteInputs.discard("newBFKernel") 

264 if config.doDefect is not True: 

265 self.prerequisiteInputs.discard("defects") 

266 if config.doDark is not True: 

267 self.prerequisiteInputs.discard("dark") 

268 if config.doFlat is not True: 

269 self.prerequisiteInputs.discard("flat") 

270 if config.doAttachTransmissionCurve is not True: 

271 self.prerequisiteInputs.discard("opticsTransmission") 

272 self.prerequisiteInputs.discard("filterTransmission") 

273 self.prerequisiteInputs.discard("sensorTransmission") 

274 self.prerequisiteInputs.discard("atmosphereTransmission") 

275 if config.doUseOpticsTransmission is not True: 

276 self.prerequisiteInputs.discard("opticsTransmission") 

277 if config.doUseFilterTransmission is not True: 

278 self.prerequisiteInputs.discard("filterTransmission") 

279 if config.doUseSensorTransmission is not True: 

280 self.prerequisiteInputs.discard("sensorTransmission") 

281 if config.doUseAtmosphereTransmission is not True: 

282 self.prerequisiteInputs.discard("atmosphereTransmission") 

283 if config.doIlluminationCorrection is not True: 

284 self.prerequisiteInputs.discard("illumMaskedImage") 

285 

286 if config.doWrite is not True: 

287 self.outputs.discard("outputExposure") 

288 self.outputs.discard("preInterpExposure") 

289 self.outputs.discard("outputFlattenedThumbnail") 

290 self.outputs.discard("outputOssThumbnail") 

291 if config.doSaveInterpPixels is not True: 

292 self.outputs.discard("preInterpExposure") 

293 if config.qa.doThumbnailOss is not True: 

294 self.outputs.discard("outputOssThumbnail") 

295 if config.qa.doThumbnailFlattened is not True: 

296 self.outputs.discard("outputFlattenedThumbnail") 

297 

298 

299class IsrTaskConfig(pipeBase.PipelineTaskConfig, 

300 pipelineConnections=IsrTaskConnections): 

301 """Configuration parameters for IsrTask. 

302 

303 Items are grouped in the order in which they are executed by the task. 

304 """ 

305 datasetType = pexConfig.Field( 

306 dtype=str, 

307 doc="Dataset type for input data; users will typically leave this alone, " 

308 "but camera-specific ISR tasks will override it", 

309 default="raw", 

310 ) 

311 

312 fallbackFilterName = pexConfig.Field( 

313 dtype=str, 

314 doc="Fallback default filter name for calibrations.", 

315 optional=True 

316 ) 

317 useFallbackDate = pexConfig.Field( 

318 dtype=bool, 

319 doc="Pass observation date when using fallback filter.", 

320 default=False, 

321 ) 

322 expectWcs = pexConfig.Field( 

323 dtype=bool, 

324 default=True, 

325 doc="Expect input science images to have a WCS (set False for e.g. spectrographs)." 

326 ) 

327 fwhm = pexConfig.Field( 

328 dtype=float, 

329 doc="FWHM of PSF in arcseconds.", 

330 default=1.0, 

331 ) 

332 qa = pexConfig.ConfigField( 

333 dtype=isrQa.IsrQaConfig, 

334 doc="QA related configuration options.", 

335 ) 

336 

337 # Image conversion configuration 

338 doConvertIntToFloat = pexConfig.Field( 

339 dtype=bool, 

340 doc="Convert integer raw images to floating point values?", 

341 default=True, 

342 ) 

343 

344 # Saturated pixel handling. 

345 doSaturation = pexConfig.Field( 

346 dtype=bool, 

347 doc="Mask saturated pixels? NB: this is totally independent of the" 

348 " interpolation option - this is ONLY setting the bits in the mask." 

349 " To have them interpolated make sure doSaturationInterpolation=True", 

350 default=True, 

351 ) 

352 saturatedMaskName = pexConfig.Field( 

353 dtype=str, 

354 doc="Name of mask plane to use in saturation detection and interpolation", 

355 default="SAT", 

356 ) 

357 saturation = pexConfig.Field( 

358 dtype=float, 

359 doc="The saturation level to use if no Detector is present in the Exposure (ignored if NaN)", 

360 default=float("NaN"), 

361 ) 

362 growSaturationFootprintSize = pexConfig.Field( 

363 dtype=int, 

364 doc="Number of pixels by which to grow the saturation footprints", 

365 default=1, 

366 ) 

367 

368 # Suspect pixel handling. 

369 doSuspect = pexConfig.Field( 

370 dtype=bool, 

371 doc="Mask suspect pixels?", 

372 default=False, 

373 ) 

374 suspectMaskName = pexConfig.Field( 

375 dtype=str, 

376 doc="Name of mask plane to use for suspect pixels", 

377 default="SUSPECT", 

378 ) 

379 numEdgeSuspect = pexConfig.Field( 

380 dtype=int, 

381 doc="Number of edge pixels to be flagged as untrustworthy.", 

382 default=0, 

383 ) 

384 edgeMaskLevel = pexConfig.ChoiceField( 

385 dtype=str, 

386 doc="Mask edge pixels in which coordinate frame: DETECTOR or AMP?", 

387 default="DETECTOR", 

388 allowed={ 

389 'DETECTOR': 'Mask only the edges of the full detector.', 

390 'AMP': 'Mask edges of each amplifier.', 

391 }, 

392 ) 

393 

394 # Initial masking options. 

395 doSetBadRegions = pexConfig.Field( 

396 dtype=bool, 

397 doc="Should we set the level of all BAD patches of the chip to the chip's average value?", 

398 default=True, 

399 ) 

400 badStatistic = pexConfig.ChoiceField( 

401 dtype=str, 

402 doc="How to estimate the average value for BAD regions.", 

403 default='MEANCLIP', 

404 allowed={ 

405 "MEANCLIP": "Correct using the (clipped) mean of good data", 

406 "MEDIAN": "Correct using the median of the good data", 

407 }, 

408 ) 

409 

410 # Overscan subtraction configuration. 

411 doOverscan = pexConfig.Field( 

412 dtype=bool, 

413 doc="Do overscan subtraction?", 

414 default=True, 

415 ) 

416 overscan = pexConfig.ConfigurableField( 

417 target=OverscanCorrectionTask, 

418 doc="Overscan subtraction task for image segments.", 

419 ) 

420 

421 overscanFitType = pexConfig.ChoiceField( 

422 dtype=str, 

423 doc="The method for fitting the overscan bias level.", 

424 default='MEDIAN', 

425 allowed={ 

426 "POLY": "Fit ordinary polynomial to the longest axis of the overscan region", 

427 "CHEB": "Fit Chebyshev polynomial to the longest axis of the overscan region", 

428 "LEG": "Fit Legendre polynomial to the longest axis of the overscan region", 

429 "NATURAL_SPLINE": "Fit natural spline to the longest axis of the overscan region", 

430 "CUBIC_SPLINE": "Fit cubic spline to the longest axis of the overscan region", 

431 "AKIMA_SPLINE": "Fit Akima spline to the longest axis of the overscan region", 

432 "MEAN": "Correct using the mean of the overscan region", 

433 "MEANCLIP": "Correct using a clipped mean of the overscan region", 

434 "MEDIAN": "Correct using the median of the overscan region", 

435 "MEDIAN_PER_ROW": "Correct using the median per row of the overscan region", 

436 }, 

437 deprecated=("Please configure overscan via the OverscanCorrectionConfig interface." 

438 " This option will no longer be used, and will be removed after v20.") 

439 ) 

440 overscanOrder = pexConfig.Field( 

441 dtype=int, 

442 doc=("Order of polynomial or to fit if overscan fit type is a polynomial, " 

443 "or number of spline knots if overscan fit type is a spline."), 

444 default=1, 

445 deprecated=("Please configure overscan via the OverscanCorrectionConfig interface." 

446 " This option will no longer be used, and will be removed after v20.") 

447 ) 

448 overscanNumSigmaClip = pexConfig.Field( 

449 dtype=float, 

450 doc="Rejection threshold (sigma) for collapsing overscan before fit", 

451 default=3.0, 

452 deprecated=("Please configure overscan via the OverscanCorrectionConfig interface." 

453 " This option will no longer be used, and will be removed after v20.") 

454 ) 

455 overscanIsInt = pexConfig.Field( 

456 dtype=bool, 

457 doc="Treat overscan as an integer image for purposes of overscan.FitType=MEDIAN" 

458 " and overscan.FitType=MEDIAN_PER_ROW.", 

459 default=True, 

460 deprecated=("Please configure overscan via the OverscanCorrectionConfig interface." 

461 " This option will no longer be used, and will be removed after v20.") 

462 ) 

463 # These options do not get deprecated, as they define how we slice up the image data. 

464 overscanNumLeadingColumnsToSkip = pexConfig.Field( 

465 dtype=int, 

466 doc="Number of columns to skip in overscan, i.e. those closest to amplifier", 

467 default=0, 

468 ) 

469 overscanNumTrailingColumnsToSkip = pexConfig.Field( 

470 dtype=int, 

471 doc="Number of columns to skip in overscan, i.e. those farthest from amplifier", 

472 default=0, 

473 ) 

474 overscanMaxDev = pexConfig.Field( 474 ↛ exitline 474 didn't jump to the function exit

475 dtype=float, 

476 doc="Maximum deviation from the median for overscan", 

477 default=1000.0, check=lambda x: x > 0 

478 ) 

479 overscanBiasJump = pexConfig.Field( 

480 dtype=bool, 

481 doc="Fit the overscan in a piecewise-fashion to correct for bias jumps?", 

482 default=False, 

483 ) 

484 overscanBiasJumpKeyword = pexConfig.Field( 

485 dtype=str, 

486 doc="Header keyword containing information about devices.", 

487 default="NO_SUCH_KEY", 

488 ) 

489 overscanBiasJumpDevices = pexConfig.ListField( 

490 dtype=str, 

491 doc="List of devices that need piecewise overscan correction.", 

492 default=(), 

493 ) 

494 overscanBiasJumpLocation = pexConfig.Field( 

495 dtype=int, 

496 doc="Location of bias jump along y-axis.", 

497 default=0, 

498 ) 

499 

500 # Amplifier to CCD assembly configuration 

501 doAssembleCcd = pexConfig.Field( 

502 dtype=bool, 

503 default=True, 

504 doc="Assemble amp-level exposures into a ccd-level exposure?" 

505 ) 

506 assembleCcd = pexConfig.ConfigurableField( 

507 target=AssembleCcdTask, 

508 doc="CCD assembly task", 

509 ) 

510 

511 # General calibration configuration. 

512 doAssembleIsrExposures = pexConfig.Field( 

513 dtype=bool, 

514 default=False, 

515 doc="Assemble amp-level calibration exposures into ccd-level exposure?" 

516 ) 

517 doTrimToMatchCalib = pexConfig.Field( 

518 dtype=bool, 

519 default=False, 

520 doc="Trim raw data to match calibration bounding boxes?" 

521 ) 

522 

523 # Bias subtraction. 

524 doBias = pexConfig.Field( 

525 dtype=bool, 

526 doc="Apply bias frame correction?", 

527 default=True, 

528 ) 

529 biasDataProductName = pexConfig.Field( 

530 dtype=str, 

531 doc="Name of the bias data product", 

532 default="bias", 

533 ) 

534 doBiasBeforeOverscan = pexConfig.Field( 

535 dtype=bool, 

536 doc="Reverse order of overscan and bias correction.", 

537 default=False 

538 ) 

539 

540 # Variance construction 

541 doVariance = pexConfig.Field( 

542 dtype=bool, 

543 doc="Calculate variance?", 

544 default=True 

545 ) 

546 gain = pexConfig.Field( 

547 dtype=float, 

548 doc="The gain to use if no Detector is present in the Exposure (ignored if NaN)", 

549 default=float("NaN"), 

550 ) 

551 readNoise = pexConfig.Field( 

552 dtype=float, 

553 doc="The read noise to use if no Detector is present in the Exposure", 

554 default=0.0, 

555 ) 

556 doEmpiricalReadNoise = pexConfig.Field( 

557 dtype=bool, 

558 default=False, 

559 doc="Calculate empirical read noise instead of value from AmpInfo data?" 

560 ) 

561 

562 # Linearization. 

563 doLinearize = pexConfig.Field( 

564 dtype=bool, 

565 doc="Correct for nonlinearity of the detector's response?", 

566 default=True, 

567 ) 

568 

569 # Crosstalk. 

570 doCrosstalk = pexConfig.Field( 

571 dtype=bool, 

572 doc="Apply intra-CCD crosstalk correction?", 

573 default=False, 

574 ) 

575 doCrosstalkBeforeAssemble = pexConfig.Field( 

576 dtype=bool, 

577 doc="Apply crosstalk correction before CCD assembly, and before trimming?", 

578 default=False, 

579 ) 

580 crosstalk = pexConfig.ConfigurableField( 

581 target=CrosstalkTask, 

582 doc="Intra-CCD crosstalk correction", 

583 ) 

584 

585 # Masking options. 

586 doDefect = pexConfig.Field( 

587 dtype=bool, 

588 doc="Apply correction for CCD defects, e.g. hot pixels?", 

589 default=True, 

590 ) 

591 doNanMasking = pexConfig.Field( 

592 dtype=bool, 

593 doc="Mask NAN pixels?", 

594 default=True, 

595 ) 

596 doWidenSaturationTrails = pexConfig.Field( 

597 dtype=bool, 

598 doc="Widen bleed trails based on their width?", 

599 default=True 

600 ) 

601 

602 # Brighter-Fatter correction. 

603 doBrighterFatter = pexConfig.Field( 

604 dtype=bool, 

605 default=False, 

606 doc="Apply the brighter fatter correction" 

607 ) 

608 brighterFatterLevel = pexConfig.ChoiceField( 

609 dtype=str, 

610 default="DETECTOR", 

611 doc="The level at which to correct for brighter-fatter.", 

612 allowed={ 

613 "AMP": "Every amplifier treated separately.", 

614 "DETECTOR": "One kernel per detector", 

615 } 

616 ) 

617 brighterFatterMaxIter = pexConfig.Field( 

618 dtype=int, 

619 default=10, 

620 doc="Maximum number of iterations for the brighter fatter correction" 

621 ) 

622 brighterFatterThreshold = pexConfig.Field( 

623 dtype=float, 

624 default=1000, 

625 doc="Threshold used to stop iterating the brighter fatter correction. It is the " 

626 " absolute value of the difference between the current corrected image and the one" 

627 " from the previous iteration summed over all the pixels." 

628 ) 

629 brighterFatterApplyGain = pexConfig.Field( 

630 dtype=bool, 

631 default=True, 

632 doc="Should the gain be applied when applying the brighter fatter correction?" 

633 ) 

634 brighterFatterMaskGrowSize = pexConfig.Field( 

635 dtype=int, 

636 default=0, 

637 doc="Number of pixels to grow the masks listed in config.maskListToInterpolate " 

638 " when brighter-fatter correction is applied." 

639 ) 

640 

641 # Dark subtraction. 

642 doDark = pexConfig.Field( 

643 dtype=bool, 

644 doc="Apply dark frame correction?", 

645 default=True, 

646 ) 

647 darkDataProductName = pexConfig.Field( 

648 dtype=str, 

649 doc="Name of the dark data product", 

650 default="dark", 

651 ) 

652 

653 # Camera-specific stray light removal. 

654 doStrayLight = pexConfig.Field( 

655 dtype=bool, 

656 doc="Subtract stray light in the y-band (due to encoder LEDs)?", 

657 default=False, 

658 ) 

659 strayLight = pexConfig.ConfigurableField( 

660 target=StrayLightTask, 

661 doc="y-band stray light correction" 

662 ) 

663 

664 # Flat correction. 

665 doFlat = pexConfig.Field( 

666 dtype=bool, 

667 doc="Apply flat field correction?", 

668 default=True, 

669 ) 

670 flatDataProductName = pexConfig.Field( 

671 dtype=str, 

672 doc="Name of the flat data product", 

673 default="flat", 

674 ) 

675 flatScalingType = pexConfig.ChoiceField( 

676 dtype=str, 

677 doc="The method for scaling the flat on the fly.", 

678 default='USER', 

679 allowed={ 

680 "USER": "Scale by flatUserScale", 

681 "MEAN": "Scale by the inverse of the mean", 

682 "MEDIAN": "Scale by the inverse of the median", 

683 }, 

684 ) 

685 flatUserScale = pexConfig.Field( 

686 dtype=float, 

687 doc="If flatScalingType is 'USER' then scale flat by this amount; ignored otherwise", 

688 default=1.0, 

689 ) 

690 doTweakFlat = pexConfig.Field( 

691 dtype=bool, 

692 doc="Tweak flats to match observed amplifier ratios?", 

693 default=False 

694 ) 

695 

696 # Amplifier normalization based on gains instead of using flats configuration. 

697 doApplyGains = pexConfig.Field( 

698 dtype=bool, 

699 doc="Correct the amplifiers for their gains instead of applying flat correction", 

700 default=False, 

701 ) 

702 normalizeGains = pexConfig.Field( 

703 dtype=bool, 

704 doc="Normalize all the amplifiers in each CCD to have the same median value.", 

705 default=False, 

706 ) 

707 

708 # Fringe correction. 

709 doFringe = pexConfig.Field( 

710 dtype=bool, 

711 doc="Apply fringe correction?", 

712 default=True, 

713 ) 

714 fringe = pexConfig.ConfigurableField( 

715 target=FringeTask, 

716 doc="Fringe subtraction task", 

717 ) 

718 fringeAfterFlat = pexConfig.Field( 

719 dtype=bool, 

720 doc="Do fringe subtraction after flat-fielding?", 

721 default=True, 

722 ) 

723 

724 # Initial CCD-level background statistics options. 

725 doMeasureBackground = pexConfig.Field( 

726 dtype=bool, 

727 doc="Measure the background level on the reduced image?", 

728 default=False, 

729 ) 

730 

731 # Camera-specific masking configuration. 

732 doCameraSpecificMasking = pexConfig.Field( 

733 dtype=bool, 

734 doc="Mask camera-specific bad regions?", 

735 default=False, 

736 ) 

737 masking = pexConfig.ConfigurableField( 

738 target=MaskingTask, 

739 doc="Masking task." 

740 ) 

741 

742 # Interpolation options. 

743 

744 doInterpolate = pexConfig.Field( 

745 dtype=bool, 

746 doc="Interpolate masked pixels?", 

747 default=True, 

748 ) 

749 doSaturationInterpolation = pexConfig.Field( 

750 dtype=bool, 

751 doc="Perform interpolation over pixels masked as saturated?" 

752 " NB: This is independent of doSaturation; if that is False this plane" 

753 " will likely be blank, resulting in a no-op here.", 

754 default=True, 

755 ) 

756 doNanInterpolation = pexConfig.Field( 

757 dtype=bool, 

758 doc="Perform interpolation over pixels masked as NaN?" 

759 " NB: This is independent of doNanMasking; if that is False this plane" 

760 " will likely be blank, resulting in a no-op here.", 

761 default=True, 

762 ) 

763 doNanInterpAfterFlat = pexConfig.Field( 

764 dtype=bool, 

765 doc=("If True, ensure we interpolate NaNs after flat-fielding, even if we " 

766 "also have to interpolate them before flat-fielding."), 

767 default=False, 

768 ) 

769 maskListToInterpolate = pexConfig.ListField( 

770 dtype=str, 

771 doc="List of mask planes that should be interpolated.", 

772 default=['SAT', 'BAD', 'UNMASKEDNAN'], 

773 ) 

774 doSaveInterpPixels = pexConfig.Field( 

775 dtype=bool, 

776 doc="Save a copy of the pre-interpolated pixel values?", 

777 default=False, 

778 ) 

779 

780 # Default photometric calibration options. 

781 fluxMag0T1 = pexConfig.DictField( 

782 keytype=str, 

783 itemtype=float, 

784 doc="The approximate flux of a zero-magnitude object in a one-second exposure, per filter.", 

785 default=dict((f, pow(10.0, 0.4*m)) for f, m in (("Unknown", 28.0), 

786 )) 

787 ) 

788 defaultFluxMag0T1 = pexConfig.Field( 

789 dtype=float, 

790 doc="Default value for fluxMag0T1 (for an unrecognized filter).", 

791 default=pow(10.0, 0.4*28.0) 

792 ) 

793 

794 # Vignette correction configuration. 

795 doVignette = pexConfig.Field( 

796 dtype=bool, 

797 doc="Apply vignetting parameters?", 

798 default=False, 

799 ) 

800 vignette = pexConfig.ConfigurableField( 

801 target=VignetteTask, 

802 doc="Vignetting task.", 

803 ) 

804 

805 # Transmission curve configuration. 

806 doAttachTransmissionCurve = pexConfig.Field( 

807 dtype=bool, 

808 default=False, 

809 doc="Construct and attach a wavelength-dependent throughput curve for this CCD image?" 

810 ) 

811 doUseOpticsTransmission = pexConfig.Field( 

812 dtype=bool, 

813 default=True, 

814 doc="Load and use transmission_optics (if doAttachTransmissionCurve is True)?" 

815 ) 

816 doUseFilterTransmission = pexConfig.Field( 

817 dtype=bool, 

818 default=True, 

819 doc="Load and use transmission_filter (if doAttachTransmissionCurve is True)?" 

820 ) 

821 doUseSensorTransmission = pexConfig.Field( 

822 dtype=bool, 

823 default=True, 

824 doc="Load and use transmission_sensor (if doAttachTransmissionCurve is True)?" 

825 ) 

826 doUseAtmosphereTransmission = pexConfig.Field( 

827 dtype=bool, 

828 default=True, 

829 doc="Load and use transmission_atmosphere (if doAttachTransmissionCurve is True)?" 

830 ) 

831 

832 # Illumination correction. 

833 doIlluminationCorrection = pexConfig.Field( 

834 dtype=bool, 

835 default=False, 

836 doc="Perform illumination correction?" 

837 ) 

838 illuminationCorrectionDataProductName = pexConfig.Field( 

839 dtype=str, 

840 doc="Name of the illumination correction data product.", 

841 default="illumcor", 

842 ) 

843 illumScale = pexConfig.Field( 

844 dtype=float, 

845 doc="Scale factor for the illumination correction.", 

846 default=1.0, 

847 ) 

848 illumFilters = pexConfig.ListField( 

849 dtype=str, 

850 default=[], 

851 doc="Only perform illumination correction for these filters." 

852 ) 

853 

854 # Write the outputs to disk. If ISR is run as a subtask, this may not be needed. 

855 doWrite = pexConfig.Field( 

856 dtype=bool, 

857 doc="Persist postISRCCD?", 

858 default=True, 

859 ) 

860 

861 def validate(self): 

862 super().validate() 

863 if self.doFlat and self.doApplyGains: 

864 raise ValueError("You may not specify both doFlat and doApplyGains") 

865 if self.doBiasBeforeOverscan and self.doTrimToMatchCalib: 

866 raise ValueError("You may not specify both doBiasBeforeOverscan and doTrimToMatchCalib") 

867 if self.doSaturationInterpolation and self.saturatedMaskName not in self.maskListToInterpolate: 

868 self.maskListToInterpolate.append(self.saturatedMaskName) 

869 if not self.doSaturationInterpolation and self.saturatedMaskName in self.maskListToInterpolate: 

870 self.maskListToInterpolate.remove(self.saturatedMaskName) 

871 if self.doNanInterpolation and "UNMASKEDNAN" not in self.maskListToInterpolate: 

872 self.maskListToInterpolate.append("UNMASKEDNAN") 

873 

874 

875class IsrTask(pipeBase.PipelineTask, pipeBase.CmdLineTask): 

876 """Apply common instrument signature correction algorithms to a raw frame. 

877 

878 The process for correcting imaging data is very similar from 

879 camera to camera. This task provides a vanilla implementation of 

880 doing these corrections, including the ability to turn certain 

881 corrections off if they are not needed. The inputs to the primary 

882 method, `run()`, are a raw exposure to be corrected and the 

883 calibration data products. The raw input is a single chip sized 

884 mosaic of all amps including overscans and other non-science 

885 pixels. The method `runDataRef()` identifies and defines the 

886 calibration data products, and is intended for use by a 

887 `lsst.pipe.base.cmdLineTask.CmdLineTask` and takes as input only a 

888 `daf.persistence.butlerSubset.ButlerDataRef`. This task may be 

889 subclassed for different camera, although the most camera specific 

890 methods have been split into subtasks that can be redirected 

891 appropriately. 

892 

893 The __init__ method sets up the subtasks for ISR processing, using 

894 the defaults from `lsst.ip.isr`. 

895 

896 Parameters 

897 ---------- 

898 args : `list` 

899 Positional arguments passed to the Task constructor. None used at this time. 

900 kwargs : `dict`, optional 

901 Keyword arguments passed on to the Task constructor. None used at this time. 

902 """ 

903 ConfigClass = IsrTaskConfig 

904 _DefaultName = "isr" 

905 

906 def __init__(self, **kwargs): 

907 super().__init__(**kwargs) 

908 self.makeSubtask("assembleCcd") 

909 self.makeSubtask("crosstalk") 

910 self.makeSubtask("strayLight") 

911 self.makeSubtask("fringe") 

912 self.makeSubtask("masking") 

913 self.makeSubtask("overscan") 

914 self.makeSubtask("vignette") 

915 

916 def runQuantum(self, butlerQC, inputRefs, outputRefs): 

917 inputs = butlerQC.get(inputRefs) 

918 

919 try: 

920 inputs['detectorNum'] = inputRefs.ccdExposure.dataId['detector'] 

921 except Exception as e: 

922 raise ValueError("Failure to find valid detectorNum value for Dataset %s: %s." % 

923 (inputRefs, e)) 

924 

925 inputs['isGen3'] = True 

926 

927 detector = inputs['ccdExposure'].getDetector() 

928 

929 if self.config.doCrosstalk is True: 

930 # Crosstalk sources need to be defined by the pipeline 

931 # yaml if they exist. 

932 if 'crosstalk' in inputs and inputs['crosstalk'] is not None: 

933 if not isinstance(inputs['crosstalk'], CrosstalkCalib): 

934 inputs['crosstalk'] = CrosstalkCalib.fromTable(inputs['crosstalk']) 

935 else: 

936 coeffVector = (self.config.crosstalk.crosstalkValues 

937 if self.config.crosstalk.useConfigCoefficients else None) 

938 crosstalkCalib = CrosstalkCalib().fromDetector(detector, coeffVector=coeffVector) 

939 inputs['crosstalk'] = crosstalkCalib 

940 if inputs['crosstalk'].interChip and len(inputs['crosstalk'].interChip) > 0: 

941 if 'crosstalkSources' not in inputs: 

942 self.log.warn("No crosstalkSources found for chip with interChip terms!") 

943 

944 if self.doLinearize(detector) is True: 

945 if 'linearizer' in inputs and isinstance(inputs['linearizer'], dict): 

946 linearizer = linearize.Linearizer(detector=detector, log=self.log) 

947 linearizer.fromYaml(inputs['linearizer']) 

948 else: 

949 linearizer = linearize.Linearizer(table=inputs.get('linearizer', None), detector=detector, 

950 log=self.log) 

951 inputs['linearizer'] = linearizer 

952 

953 if self.config.doDefect is True: 

954 if "defects" in inputs and inputs['defects'] is not None: 

955 # defects is loaded as a BaseCatalog with columns x0, y0, width, height. 

956 # masking expects a list of defects defined by their bounding box 

957 if not isinstance(inputs["defects"], Defects): 

958 inputs["defects"] = Defects.fromTable(inputs["defects"]) 

959 

960 # Load the correct style of brighter fatter kernel, and repack 

961 # the information as a numpy array. 

962 if self.config.doBrighterFatter: 

963 brighterFatterKernel = inputs.pop('newBFKernel', None) 

964 if brighterFatterKernel is None: 

965 brighterFatterKernel = inputs.get('bfKernel', None) 

966 

967 if brighterFatterKernel is not None and not isinstance(brighterFatterKernel, numpy.ndarray): 

968 detId = detector.getId() 

969 inputs['bfGains'] = brighterFatterKernel.gain 

970 # If the kernel is not an ndarray, it's the cp_pipe version 

971 # so extract the kernel for this detector, or raise an error 

972 if self.config.brighterFatterLevel == 'DETECTOR': 

973 if brighterFatterKernel.detectorKernel: 

974 inputs['bfKernel'] = brighterFatterKernel.detectorKernel[detId] 

975 elif brighterFatterKernel.detectorKernelFromAmpKernels: 

976 inputs['bfKernel'] = brighterFatterKernel.detectorKernelFromAmpKernels[detId] 

977 else: 

978 raise RuntimeError("Failed to extract kernel from new-style BF kernel.") 

979 else: 

980 # TODO DM-15631 for implementing this 

981 raise NotImplementedError("Per-amplifier brighter-fatter correction not implemented") 

982 

983 if self.config.doFringe is True and self.fringe.checkFilter(inputs['ccdExposure']): 

984 expId = inputs['ccdExposure'].getInfo().getVisitInfo().getExposureId() 

985 inputs['fringes'] = self.fringe.loadFringes(inputs['fringes'], 

986 expId=expId, 

987 assembler=self.assembleCcd 

988 if self.config.doAssembleIsrExposures else None) 

989 else: 

990 inputs['fringes'] = pipeBase.Struct(fringes=None) 

991 

992 if self.config.doStrayLight is True and self.strayLight.checkFilter(inputs['ccdExposure']): 

993 if 'strayLightData' not in inputs: 

994 inputs['strayLightData'] = None 

995 

996 outputs = self.run(**inputs) 

997 butlerQC.put(outputs, outputRefs) 

998 

999 def readIsrData(self, dataRef, rawExposure): 

1000 """Retrieve necessary frames for instrument signature removal. 

1001 

1002 Pre-fetching all required ISR data products limits the IO 

1003 required by the ISR. Any conflict between the calibration data 

1004 available and that needed for ISR is also detected prior to 

1005 doing processing, allowing it to fail quickly. 

1006 

1007 Parameters 

1008 ---------- 

1009 dataRef : `daf.persistence.butlerSubset.ButlerDataRef` 

1010 Butler reference of the detector data to be processed 

1011 rawExposure : `afw.image.Exposure` 

1012 The raw exposure that will later be corrected with the 

1013 retrieved calibration data; should not be modified in this 

1014 method. 

1015 

1016 Returns 

1017 ------- 

1018 result : `lsst.pipe.base.Struct` 

1019 Result struct with components (which may be `None`): 

1020 - ``bias``: bias calibration frame (`afw.image.Exposure`) 

1021 - ``linearizer``: functor for linearization (`ip.isr.linearize.LinearizeBase`) 

1022 - ``crosstalkSources``: list of possible crosstalk sources (`list`) 

1023 - ``dark``: dark calibration frame (`afw.image.Exposure`) 

1024 - ``flat``: flat calibration frame (`afw.image.Exposure`) 

1025 - ``bfKernel``: Brighter-Fatter kernel (`numpy.ndarray`) 

1026 - ``defects``: list of defects (`lsst.ip.isr.Defects`) 

1027 - ``fringes``: `lsst.pipe.base.Struct` with components: 

1028 - ``fringes``: fringe calibration frame (`afw.image.Exposure`) 

1029 - ``seed``: random seed derived from the ccdExposureId for random 

1030 number generator (`uint32`). 

1031 - ``opticsTransmission``: `lsst.afw.image.TransmissionCurve` 

1032 A ``TransmissionCurve`` that represents the throughput of the optics, 

1033 to be evaluated in focal-plane coordinates. 

1034 - ``filterTransmission`` : `lsst.afw.image.TransmissionCurve` 

1035 A ``TransmissionCurve`` that represents the throughput of the filter 

1036 itself, to be evaluated in focal-plane coordinates. 

1037 - ``sensorTransmission`` : `lsst.afw.image.TransmissionCurve` 

1038 A ``TransmissionCurve`` that represents the throughput of the sensor 

1039 itself, to be evaluated in post-assembly trimmed detector coordinates. 

1040 - ``atmosphereTransmission`` : `lsst.afw.image.TransmissionCurve` 

1041 A ``TransmissionCurve`` that represents the throughput of the 

1042 atmosphere, assumed to be spatially constant. 

1043 - ``strayLightData`` : `object` 

1044 An opaque object containing calibration information for 

1045 stray-light correction. If `None`, no correction will be 

1046 performed. 

1047 - ``illumMaskedImage`` : illumination correction image (`lsst.afw.image.MaskedImage`) 

1048 

1049 Raises 

1050 ------ 

1051 NotImplementedError : 

1052 Raised if a per-amplifier brighter-fatter kernel is requested by the configuration. 

1053 """ 

1054 try: 

1055 dateObs = rawExposure.getInfo().getVisitInfo().getDate() 

1056 dateObs = dateObs.toPython().isoformat() 

1057 except RuntimeError: 

1058 self.log.warn("Unable to identify dateObs for rawExposure.") 

1059 dateObs = None 

1060 

1061 ccd = rawExposure.getDetector() 

1062 filterName = afwImage.Filter(rawExposure.getFilter().getId()).getName() # Canonical name for filter 

1063 rawExposure.mask.addMaskPlane("UNMASKEDNAN") # needed to match pre DM-15862 processing. 

1064 biasExposure = (self.getIsrExposure(dataRef, self.config.biasDataProductName) 

1065 if self.config.doBias else None) 

1066 # immediate=True required for functors and linearizers are functors; see ticket DM-6515 

1067 linearizer = (dataRef.get("linearizer", immediate=True) 

1068 if self.doLinearize(ccd) else None) 

1069 if linearizer is not None and not isinstance(linearizer, numpy.ndarray): 

1070 linearizer.log = self.log 

1071 if isinstance(linearizer, numpy.ndarray): 

1072 linearizer = linearize.Linearizer(table=linearizer, detector=ccd) 

1073 

1074 crosstalkCalib = None 

1075 if self.config.doCrosstalk: 

1076 try: 

1077 crosstalkCalib = dataRef.get("crosstalk", immediate=True) 

1078 except NoResults: 

1079 coeffVector = (self.config.crosstalk.crosstalkValues 

1080 if self.config.crosstalk.useConfigCoefficients else None) 

1081 crosstalkCalib = CrosstalkCalib().fromDetector(ccd, coeffVector=coeffVector) 

1082 crosstalkSources = (self.crosstalk.prepCrosstalk(dataRef, crosstalkCalib) 

1083 if self.config.doCrosstalk else None) 

1084 

1085 darkExposure = (self.getIsrExposure(dataRef, self.config.darkDataProductName) 

1086 if self.config.doDark else None) 

1087 flatExposure = (self.getIsrExposure(dataRef, self.config.flatDataProductName, 

1088 dateObs=dateObs) 

1089 if self.config.doFlat else None) 

1090 

1091 brighterFatterKernel = None 

1092 brighterFatterGains = None 

1093 if self.config.doBrighterFatter is True: 

1094 try: 

1095 # Use the new-style cp_pipe version of the kernel if it exists 

1096 # If using a new-style kernel, always use the self-consistent 

1097 # gains, i.e. the ones inside the kernel object itself 

1098 brighterFatterKernel = dataRef.get("brighterFatterKernel") 

1099 brighterFatterGains = brighterFatterKernel.gain 

1100 self.log.info("New style bright-fatter kernel (brighterFatterKernel) loaded") 

1101 except NoResults: 

1102 try: # Fall back to the old-style numpy-ndarray style kernel if necessary. 

1103 brighterFatterKernel = dataRef.get("bfKernel") 

1104 self.log.info("Old style bright-fatter kernel (np.array) loaded") 

1105 except NoResults: 

1106 brighterFatterKernel = None 

1107 if brighterFatterKernel is not None and not isinstance(brighterFatterKernel, numpy.ndarray): 

1108 # If the kernel is not an ndarray, it's the cp_pipe version 

1109 # so extract the kernel for this detector, or raise an error 

1110 if self.config.brighterFatterLevel == 'DETECTOR': 

1111 if brighterFatterKernel.detectorKernel: 

1112 brighterFatterKernel = brighterFatterKernel.detectorKernel[ccd.getId()] 

1113 elif brighterFatterKernel.detectorKernelFromAmpKernels: 

1114 brighterFatterKernel = brighterFatterKernel.detectorKernelFromAmpKernels[ccd.getId()] 

1115 else: 

1116 raise RuntimeError("Failed to extract kernel from new-style BF kernel.") 

1117 else: 

1118 # TODO DM-15631 for implementing this 

1119 raise NotImplementedError("Per-amplifier brighter-fatter correction not implemented") 

1120 

1121 defectList = (dataRef.get("defects") 

1122 if self.config.doDefect else None) 

1123 fringeStruct = (self.fringe.readFringes(dataRef, assembler=self.assembleCcd 

1124 if self.config.doAssembleIsrExposures else None) 

1125 if self.config.doFringe and self.fringe.checkFilter(rawExposure) 

1126 else pipeBase.Struct(fringes=None)) 

1127 

1128 if self.config.doAttachTransmissionCurve: 

1129 opticsTransmission = (dataRef.get("transmission_optics") 

1130 if self.config.doUseOpticsTransmission else None) 

1131 filterTransmission = (dataRef.get("transmission_filter") 

1132 if self.config.doUseFilterTransmission else None) 

1133 sensorTransmission = (dataRef.get("transmission_sensor") 

1134 if self.config.doUseSensorTransmission else None) 

1135 atmosphereTransmission = (dataRef.get("transmission_atmosphere") 

1136 if self.config.doUseAtmosphereTransmission else None) 

1137 else: 

1138 opticsTransmission = None 

1139 filterTransmission = None 

1140 sensorTransmission = None 

1141 atmosphereTransmission = None 

1142 

1143 if self.config.doStrayLight: 

1144 strayLightData = self.strayLight.readIsrData(dataRef, rawExposure) 

1145 else: 

1146 strayLightData = None 

1147 

1148 illumMaskedImage = (self.getIsrExposure(dataRef, 

1149 self.config.illuminationCorrectionDataProductName).getMaskedImage() 

1150 if (self.config.doIlluminationCorrection 

1151 and filterName in self.config.illumFilters) 

1152 else None) 

1153 

1154 # Struct should include only kwargs to run() 

1155 return pipeBase.Struct(bias=biasExposure, 

1156 linearizer=linearizer, 

1157 crosstalk=crosstalkCalib, 

1158 crosstalkSources=crosstalkSources, 

1159 dark=darkExposure, 

1160 flat=flatExposure, 

1161 bfKernel=brighterFatterKernel, 

1162 bfGains=brighterFatterGains, 

1163 defects=defectList, 

1164 fringes=fringeStruct, 

1165 opticsTransmission=opticsTransmission, 

1166 filterTransmission=filterTransmission, 

1167 sensorTransmission=sensorTransmission, 

1168 atmosphereTransmission=atmosphereTransmission, 

1169 strayLightData=strayLightData, 

1170 illumMaskedImage=illumMaskedImage 

1171 ) 

1172 

1173 @pipeBase.timeMethod 

1174 def run(self, ccdExposure, camera=None, bias=None, linearizer=None, 

1175 crosstalk=None, crosstalkSources=None, 

1176 dark=None, flat=None, bfKernel=None, bfGains=None, defects=None, 

1177 fringes=pipeBase.Struct(fringes=None), opticsTransmission=None, filterTransmission=None, 

1178 sensorTransmission=None, atmosphereTransmission=None, 

1179 detectorNum=None, strayLightData=None, illumMaskedImage=None, 

1180 isGen3=False, 

1181 ): 

1182 """Perform instrument signature removal on an exposure. 

1183 

1184 Steps included in the ISR processing, in order performed, are: 

1185 - saturation and suspect pixel masking 

1186 - overscan subtraction 

1187 - CCD assembly of individual amplifiers 

1188 - bias subtraction 

1189 - variance image construction 

1190 - linearization of non-linear response 

1191 - crosstalk masking 

1192 - brighter-fatter correction 

1193 - dark subtraction 

1194 - fringe correction 

1195 - stray light subtraction 

1196 - flat correction 

1197 - masking of known defects and camera specific features 

1198 - vignette calculation 

1199 - appending transmission curve and distortion model 

1200 

1201 Parameters 

1202 ---------- 

1203 ccdExposure : `lsst.afw.image.Exposure` 

1204 The raw exposure that is to be run through ISR. The 

1205 exposure is modified by this method. 

1206 camera : `lsst.afw.cameraGeom.Camera`, optional 

1207 The camera geometry for this exposure. Required if ``isGen3`` is 

1208 `True` and one or more of ``ccdExposure``, ``bias``, ``dark``, or 

1209 ``flat`` does not have an associated detector. 

1210 bias : `lsst.afw.image.Exposure`, optional 

1211 Bias calibration frame. 

1212 linearizer : `lsst.ip.isr.linearize.LinearizeBase`, optional 

1213 Functor for linearization. 

1214 crosstalk : `lsst.ip.isr.crosstalk.CrosstalkCalib`, optional 

1215 Calibration for crosstalk. 

1216 crosstalkSources : `list`, optional 

1217 List of possible crosstalk sources. 

1218 dark : `lsst.afw.image.Exposure`, optional 

1219 Dark calibration frame. 

1220 flat : `lsst.afw.image.Exposure`, optional 

1221 Flat calibration frame. 

1222 bfKernel : `numpy.ndarray`, optional 

1223 Brighter-fatter kernel. 

1224 bfGains : `dict` of `float`, optional 

1225 Gains used to override the detector's nominal gains for the 

1226 brighter-fatter correction. A dict keyed by amplifier name for 

1227 the detector in question. 

1228 defects : `lsst.ip.isr.Defects`, optional 

1229 List of defects. 

1230 fringes : `lsst.pipe.base.Struct`, optional 

1231 Struct containing the fringe correction data, with 

1232 elements: 

1233 - ``fringes``: fringe calibration frame (`afw.image.Exposure`) 

1234 - ``seed``: random seed derived from the ccdExposureId for random 

1235 number generator (`uint32`) 

1236 opticsTransmission: `lsst.afw.image.TransmissionCurve`, optional 

1237 A ``TransmissionCurve`` that represents the throughput of the optics, 

1238 to be evaluated in focal-plane coordinates. 

1239 filterTransmission : `lsst.afw.image.TransmissionCurve` 

1240 A ``TransmissionCurve`` that represents the throughput of the filter 

1241 itself, to be evaluated in focal-plane coordinates. 

1242 sensorTransmission : `lsst.afw.image.TransmissionCurve` 

1243 A ``TransmissionCurve`` that represents the throughput of the sensor 

1244 itself, to be evaluated in post-assembly trimmed detector coordinates. 

1245 atmosphereTransmission : `lsst.afw.image.TransmissionCurve` 

1246 A ``TransmissionCurve`` that represents the throughput of the 

1247 atmosphere, assumed to be spatially constant. 

1248 detectorNum : `int`, optional 

1249 The integer number for the detector to process. 

1250 isGen3 : bool, optional 

1251 Flag this call to run() as using the Gen3 butler environment. 

1252 strayLightData : `object`, optional 

1253 Opaque object containing calibration information for stray-light 

1254 correction. If `None`, no correction will be performed. 

1255 illumMaskedImage : `lsst.afw.image.MaskedImage`, optional 

1256 Illumination correction image. 

1257 

1258 Returns 

1259 ------- 

1260 result : `lsst.pipe.base.Struct` 

1261 Result struct with component: 

1262 - ``exposure`` : `afw.image.Exposure` 

1263 The fully ISR corrected exposure. 

1264 - ``outputExposure`` : `afw.image.Exposure` 

1265 An alias for `exposure` 

1266 - ``ossThumb`` : `numpy.ndarray` 

1267 Thumbnail image of the exposure after overscan subtraction. 

1268 - ``flattenedThumb`` : `numpy.ndarray` 

1269 Thumbnail image of the exposure after flat-field correction. 

1270 

1271 Raises 

1272 ------ 

1273 RuntimeError 

1274 Raised if a configuration option is set to True, but the 

1275 required calibration data has not been specified. 

1276 

1277 Notes 

1278 ----- 

1279 The current processed exposure can be viewed by setting the 

1280 appropriate lsstDebug entries in the `debug.display` 

1281 dictionary. The names of these entries correspond to some of 

1282 the IsrTaskConfig Boolean options, with the value denoting the 

1283 frame to use. The exposure is shown inside the matching 

1284 option check and after the processing of that step has 

1285 finished. The steps with debug points are: 

1286 

1287 doAssembleCcd 

1288 doBias 

1289 doCrosstalk 

1290 doBrighterFatter 

1291 doDark 

1292 doFringe 

1293 doStrayLight 

1294 doFlat 

1295 

1296 In addition, setting the "postISRCCD" entry displays the 

1297 exposure after all ISR processing has finished. 

1298 

1299 """ 

1300 

1301 if isGen3 is True: 

1302 # Gen3 currently cannot automatically do configuration overrides. 

1303 # DM-15257 looks to discuss this issue. 

1304 # Configure input exposures; 

1305 if detectorNum is None: 

1306 raise RuntimeError("Must supply the detectorNum if running as Gen3.") 

1307 

1308 ccdExposure = self.ensureExposure(ccdExposure, camera, detectorNum) 

1309 bias = self.ensureExposure(bias, camera, detectorNum) 

1310 dark = self.ensureExposure(dark, camera, detectorNum) 

1311 flat = self.ensureExposure(flat, camera, detectorNum) 

1312 else: 

1313 if isinstance(ccdExposure, ButlerDataRef): 

1314 return self.runDataRef(ccdExposure) 

1315 

1316 ccd = ccdExposure.getDetector() 

1317 filterName = afwImage.Filter(ccdExposure.getFilter().getId()).getName() # Canonical name for filter 

1318 

1319 if not ccd: 

1320 assert not self.config.doAssembleCcd, "You need a Detector to run assembleCcd." 

1321 ccd = [FakeAmp(ccdExposure, self.config)] 

1322 

1323 # Validate Input 

1324 if self.config.doBias and bias is None: 

1325 raise RuntimeError("Must supply a bias exposure if config.doBias=True.") 

1326 if self.doLinearize(ccd) and linearizer is None: 

1327 raise RuntimeError("Must supply a linearizer if config.doLinearize=True for this detector.") 

1328 if self.config.doBrighterFatter and bfKernel is None: 

1329 raise RuntimeError("Must supply a kernel if config.doBrighterFatter=True.") 

1330 if self.config.doDark and dark is None: 

1331 raise RuntimeError("Must supply a dark exposure if config.doDark=True.") 

1332 if self.config.doFlat and flat is None: 

1333 raise RuntimeError("Must supply a flat exposure if config.doFlat=True.") 

1334 if self.config.doDefect and defects is None: 

1335 raise RuntimeError("Must supply defects if config.doDefect=True.") 

1336 if (self.config.doFringe and filterName in self.fringe.config.filters 

1337 and fringes.fringes is None): 

1338 # The `fringes` object needs to be a pipeBase.Struct, as 

1339 # we use it as a `dict` for the parameters of 

1340 # `FringeTask.run()`. The `fringes.fringes` `list` may 

1341 # not be `None` if `doFringe=True`. Otherwise, raise. 

1342 raise RuntimeError("Must supply fringe exposure as a pipeBase.Struct.") 

1343 if (self.config.doIlluminationCorrection and filterName in self.config.illumFilters 

1344 and illumMaskedImage is None): 

1345 raise RuntimeError("Must supply an illumcor if config.doIlluminationCorrection=True.") 

1346 

1347 # Begin ISR processing. 

1348 if self.config.doConvertIntToFloat: 

1349 self.log.info("Converting exposure to floating point values.") 

1350 ccdExposure = self.convertIntToFloat(ccdExposure) 

1351 

1352 if self.config.doBias and self.config.doBiasBeforeOverscan: 

1353 self.log.info("Applying bias correction.") 

1354 isrFunctions.biasCorrection(ccdExposure.getMaskedImage(), bias.getMaskedImage(), 

1355 trimToFit=self.config.doTrimToMatchCalib) 

1356 self.debugView(ccdExposure, "doBias") 

1357 

1358 # Amplifier level processing. 

1359 overscans = [] 

1360 for amp in ccd: 

1361 # if ccdExposure is one amp, check for coverage to prevent performing ops multiple times 

1362 if ccdExposure.getBBox().contains(amp.getBBox()): 

1363 # Check for fully masked bad amplifiers, and generate masks for SUSPECT and SATURATED values. 

1364 badAmp = self.maskAmplifier(ccdExposure, amp, defects) 

1365 

1366 if self.config.doOverscan and not badAmp: 

1367 # Overscan correction on amp-by-amp basis. 

1368 overscanResults = self.overscanCorrection(ccdExposure, amp) 

1369 self.log.debug("Corrected overscan for amplifier %s.", amp.getName()) 

1370 if overscanResults is not None and \ 

1371 self.config.qa is not None and self.config.qa.saveStats is True: 

1372 if isinstance(overscanResults.overscanFit, float): 

1373 qaMedian = overscanResults.overscanFit 

1374 qaStdev = float("NaN") 

1375 else: 

1376 qaStats = afwMath.makeStatistics(overscanResults.overscanFit, 

1377 afwMath.MEDIAN | afwMath.STDEVCLIP) 

1378 qaMedian = qaStats.getValue(afwMath.MEDIAN) 

1379 qaStdev = qaStats.getValue(afwMath.STDEVCLIP) 

1380 

1381 self.metadata.set(f"FIT MEDIAN {amp.getName()}", qaMedian) 

1382 self.metadata.set(f"FIT STDEV {amp.getName()}", qaStdev) 

1383 self.log.debug(" Overscan stats for amplifer %s: %f +/- %f", 

1384 amp.getName(), qaMedian, qaStdev) 

1385 

1386 # Residuals after overscan correction 

1387 qaStatsAfter = afwMath.makeStatistics(overscanResults.overscanImage, 

1388 afwMath.MEDIAN | afwMath.STDEVCLIP) 

1389 qaMedianAfter = qaStatsAfter.getValue(afwMath.MEDIAN) 

1390 qaStdevAfter = qaStatsAfter.getValue(afwMath.STDEVCLIP) 

1391 

1392 self.metadata.set(f"RESIDUAL MEDIAN {amp.getName()}", qaMedianAfter) 

1393 self.metadata.set(f"RESIDUAL STDEV {amp.getName()}", qaStdevAfter) 

1394 self.log.debug(" Overscan stats for amplifer %s after correction: %f +/- %f", 

1395 amp.getName(), qaMedianAfter, qaStdevAfter) 

1396 

1397 ccdExposure.getMetadata().set('OVERSCAN', "Overscan corrected") 

1398 else: 

1399 if badAmp: 

1400 self.log.warn("Amplifier %s is bad.", amp.getName()) 

1401 overscanResults = None 

1402 

1403 overscans.append(overscanResults if overscanResults is not None else None) 

1404 else: 

1405 self.log.info("Skipped OSCAN for %s.", amp.getName()) 

1406 

1407 if self.config.doCrosstalk and self.config.doCrosstalkBeforeAssemble: 

1408 self.log.info("Applying crosstalk correction.") 

1409 self.crosstalk.run(ccdExposure, crosstalk=crosstalk, 

1410 crosstalkSources=crosstalkSources) 

1411 self.debugView(ccdExposure, "doCrosstalk") 

1412 

1413 if self.config.doAssembleCcd: 

1414 self.log.info("Assembling CCD from amplifiers.") 

1415 ccdExposure = self.assembleCcd.assembleCcd(ccdExposure) 

1416 

1417 if self.config.expectWcs and not ccdExposure.getWcs(): 

1418 self.log.warn("No WCS found in input exposure.") 

1419 self.debugView(ccdExposure, "doAssembleCcd") 

1420 

1421 ossThumb = None 

1422 if self.config.qa.doThumbnailOss: 

1423 ossThumb = isrQa.makeThumbnail(ccdExposure, isrQaConfig=self.config.qa) 

1424 

1425 if self.config.doBias and not self.config.doBiasBeforeOverscan: 

1426 self.log.info("Applying bias correction.") 

1427 isrFunctions.biasCorrection(ccdExposure.getMaskedImage(), bias.getMaskedImage(), 

1428 trimToFit=self.config.doTrimToMatchCalib) 

1429 self.debugView(ccdExposure, "doBias") 

1430 

1431 if self.config.doVariance: 

1432 for amp, overscanResults in zip(ccd, overscans): 

1433 if ccdExposure.getBBox().contains(amp.getBBox()): 

1434 self.log.debug("Constructing variance map for amplifer %s.", amp.getName()) 

1435 ampExposure = ccdExposure.Factory(ccdExposure, amp.getBBox()) 

1436 if overscanResults is not None: 

1437 self.updateVariance(ampExposure, amp, 

1438 overscanImage=overscanResults.overscanImage) 

1439 else: 

1440 self.updateVariance(ampExposure, amp, 

1441 overscanImage=None) 

1442 if self.config.qa is not None and self.config.qa.saveStats is True: 

1443 qaStats = afwMath.makeStatistics(ampExposure.getVariance(), 

1444 afwMath.MEDIAN | afwMath.STDEVCLIP) 

1445 self.metadata.set(f"ISR VARIANCE {amp.getName()} MEDIAN", 

1446 qaStats.getValue(afwMath.MEDIAN)) 

1447 self.metadata.set(f"ISR VARIANCE {amp.getName()} STDEV", 

1448 qaStats.getValue(afwMath.STDEVCLIP)) 

1449 self.log.debug(" Variance stats for amplifer %s: %f +/- %f.", 

1450 amp.getName(), qaStats.getValue(afwMath.MEDIAN), 

1451 qaStats.getValue(afwMath.STDEVCLIP)) 

1452 

1453 if self.doLinearize(ccd): 

1454 self.log.info("Applying linearizer.") 

1455 linearizer.applyLinearity(image=ccdExposure.getMaskedImage().getImage(), 

1456 detector=ccd, log=self.log) 

1457 

1458 if self.config.doCrosstalk and not self.config.doCrosstalkBeforeAssemble: 

1459 self.log.info("Applying crosstalk correction.") 

1460 self.crosstalk.run(ccdExposure, crosstalk=crosstalk, 

1461 crosstalkSources=crosstalkSources, isTrimmed=True) 

1462 self.debugView(ccdExposure, "doCrosstalk") 

1463 

1464 # Masking block. Optionally mask known defects, NAN pixels, widen trails, and do 

1465 # anything else the camera needs. Saturated and suspect pixels have already been masked. 

1466 if self.config.doDefect: 

1467 self.log.info("Masking defects.") 

1468 self.maskDefect(ccdExposure, defects) 

1469 

1470 if self.config.numEdgeSuspect > 0: 

1471 self.log.info("Masking edges as SUSPECT.") 

1472 self.maskEdges(ccdExposure, numEdgePixels=self.config.numEdgeSuspect, 

1473 maskPlane="SUSPECT", level=self.config.edgeMaskLevel) 

1474 

1475 if self.config.doNanMasking: 

1476 self.log.info("Masking NAN value pixels.") 

1477 self.maskNan(ccdExposure) 

1478 

1479 if self.config.doWidenSaturationTrails: 

1480 self.log.info("Widening saturation trails.") 

1481 isrFunctions.widenSaturationTrails(ccdExposure.getMaskedImage().getMask()) 

1482 

1483 if self.config.doCameraSpecificMasking: 

1484 self.log.info("Masking regions for camera specific reasons.") 

1485 self.masking.run(ccdExposure) 

1486 

1487 if self.config.doBrighterFatter: 

1488 # We need to apply flats and darks before we can interpolate, and we 

1489 # need to interpolate before we do B-F, but we do B-F without the 

1490 # flats and darks applied so we can work in units of electrons or holes. 

1491 # This context manager applies and then removes the darks and flats. 

1492 # 

1493 # We also do not want to interpolate values here, so operate on temporary 

1494 # images so we can apply only the BF-correction and roll back the 

1495 # interpolation. 

1496 interpExp = ccdExposure.clone() 

1497 with self.flatContext(interpExp, flat, dark): 

1498 isrFunctions.interpolateFromMask( 

1499 maskedImage=interpExp.getMaskedImage(), 

1500 fwhm=self.config.fwhm, 

1501 growSaturatedFootprints=self.config.growSaturationFootprintSize, 

1502 maskNameList=self.config.maskListToInterpolate 

1503 ) 

1504 bfExp = interpExp.clone() 

1505 

1506 self.log.info("Applying brighter fatter correction using kernel type %s / gains %s.", 

1507 type(bfKernel), type(bfGains)) 

1508 bfResults = isrFunctions.brighterFatterCorrection(bfExp, bfKernel, 

1509 self.config.brighterFatterMaxIter, 

1510 self.config.brighterFatterThreshold, 

1511 self.config.brighterFatterApplyGain, 

1512 bfGains) 

1513 if bfResults[1] == self.config.brighterFatterMaxIter: 

1514 self.log.warn("Brighter fatter correction did not converge, final difference %f.", 

1515 bfResults[0]) 

1516 else: 

1517 self.log.info("Finished brighter fatter correction in %d iterations.", 

1518 bfResults[1]) 

1519 image = ccdExposure.getMaskedImage().getImage() 

1520 bfCorr = bfExp.getMaskedImage().getImage() 

1521 bfCorr -= interpExp.getMaskedImage().getImage() 

1522 image += bfCorr 

1523 

1524 # Applying the brighter-fatter correction applies a 

1525 # convolution to the science image. At the edges this 

1526 # convolution may not have sufficient valid pixels to 

1527 # produce a valid correction. Mark pixels within the size 

1528 # of the brighter-fatter kernel as EDGE to warn of this 

1529 # fact. 

1530 self.log.info("Ensuring image edges are masked as SUSPECT to the brighter-fatter kernel size.") 

1531 self.maskEdges(ccdExposure, numEdgePixels=numpy.max(bfKernel.shape) // 2, 

1532 maskPlane="EDGE") 

1533 

1534 if self.config.brighterFatterMaskGrowSize > 0: 

1535 self.log.info("Growing masks to account for brighter-fatter kernel convolution.") 

1536 for maskPlane in self.config.maskListToInterpolate: 

1537 isrFunctions.growMasks(ccdExposure.getMask(), 

1538 radius=self.config.brighterFatterMaskGrowSize, 

1539 maskNameList=maskPlane, 

1540 maskValue=maskPlane) 

1541 

1542 self.debugView(ccdExposure, "doBrighterFatter") 

1543 

1544 if self.config.doDark: 

1545 self.log.info("Applying dark correction.") 

1546 self.darkCorrection(ccdExposure, dark) 

1547 self.debugView(ccdExposure, "doDark") 

1548 

1549 if self.config.doFringe and not self.config.fringeAfterFlat: 

1550 self.log.info("Applying fringe correction before flat.") 

1551 self.fringe.run(ccdExposure, **fringes.getDict()) 

1552 self.debugView(ccdExposure, "doFringe") 

1553 

1554 if self.config.doStrayLight and self.strayLight.check(ccdExposure): 

1555 self.log.info("Checking strayLight correction.") 

1556 self.strayLight.run(ccdExposure, strayLightData) 

1557 self.debugView(ccdExposure, "doStrayLight") 

1558 

1559 if self.config.doFlat: 

1560 self.log.info("Applying flat correction.") 

1561 self.flatCorrection(ccdExposure, flat) 

1562 self.debugView(ccdExposure, "doFlat") 

1563 

1564 if self.config.doApplyGains: 

1565 self.log.info("Applying gain correction instead of flat.") 

1566 isrFunctions.applyGains(ccdExposure, self.config.normalizeGains) 

1567 

1568 if self.config.doFringe and self.config.fringeAfterFlat: 

1569 self.log.info("Applying fringe correction after flat.") 

1570 self.fringe.run(ccdExposure, **fringes.getDict()) 

1571 

1572 if self.config.doVignette: 

1573 self.log.info("Constructing Vignette polygon.") 

1574 self.vignettePolygon = self.vignette.run(ccdExposure) 

1575 

1576 if self.config.vignette.doWriteVignettePolygon: 

1577 self.setValidPolygonIntersect(ccdExposure, self.vignettePolygon) 

1578 

1579 if self.config.doAttachTransmissionCurve: 

1580 self.log.info("Adding transmission curves.") 

1581 isrFunctions.attachTransmissionCurve(ccdExposure, opticsTransmission=opticsTransmission, 

1582 filterTransmission=filterTransmission, 

1583 sensorTransmission=sensorTransmission, 

1584 atmosphereTransmission=atmosphereTransmission) 

1585 

1586 flattenedThumb = None 

1587 if self.config.qa.doThumbnailFlattened: 

1588 flattenedThumb = isrQa.makeThumbnail(ccdExposure, isrQaConfig=self.config.qa) 

1589 

1590 if self.config.doIlluminationCorrection and filterName in self.config.illumFilters: 

1591 self.log.info("Performing illumination correction.") 

1592 isrFunctions.illuminationCorrection(ccdExposure.getMaskedImage(), 

1593 illumMaskedImage, illumScale=self.config.illumScale, 

1594 trimToFit=self.config.doTrimToMatchCalib) 

1595 

1596 preInterpExp = None 

1597 if self.config.doSaveInterpPixels: 

1598 preInterpExp = ccdExposure.clone() 

1599 

1600 # Reset and interpolate bad pixels. 

1601 # 

1602 # Large contiguous bad regions (which should have the BAD mask 

1603 # bit set) should have their values set to the image median. 

1604 # This group should include defects and bad amplifiers. As the 

1605 # area covered by these defects are large, there's little 

1606 # reason to expect that interpolation would provide a more 

1607 # useful value. 

1608 # 

1609 # Smaller defects can be safely interpolated after the larger 

1610 # regions have had their pixel values reset. This ensures 

1611 # that the remaining defects adjacent to bad amplifiers (as an 

1612 # example) do not attempt to interpolate extreme values. 

1613 if self.config.doSetBadRegions: 

1614 badPixelCount, badPixelValue = isrFunctions.setBadRegions(ccdExposure) 

1615 if badPixelCount > 0: 

1616 self.log.info("Set %d BAD pixels to %f.", badPixelCount, badPixelValue) 

1617 

1618 if self.config.doInterpolate: 

1619 self.log.info("Interpolating masked pixels.") 

1620 isrFunctions.interpolateFromMask( 

1621 maskedImage=ccdExposure.getMaskedImage(), 

1622 fwhm=self.config.fwhm, 

1623 growSaturatedFootprints=self.config.growSaturationFootprintSize, 

1624 maskNameList=list(self.config.maskListToInterpolate) 

1625 ) 

1626 

1627 self.roughZeroPoint(ccdExposure) 

1628 

1629 if self.config.doMeasureBackground: 

1630 self.log.info("Measuring background level.") 

1631 self.measureBackground(ccdExposure, self.config.qa) 

1632 

1633 if self.config.qa is not None and self.config.qa.saveStats is True: 

1634 for amp in ccd: 

1635 ampExposure = ccdExposure.Factory(ccdExposure, amp.getBBox()) 

1636 qaStats = afwMath.makeStatistics(ampExposure.getImage(), 

1637 afwMath.MEDIAN | afwMath.STDEVCLIP) 

1638 self.metadata.set("ISR BACKGROUND {} MEDIAN".format(amp.getName()), 

1639 qaStats.getValue(afwMath.MEDIAN)) 

1640 self.metadata.set("ISR BACKGROUND {} STDEV".format(amp.getName()), 

1641 qaStats.getValue(afwMath.STDEVCLIP)) 

1642 self.log.debug(" Background stats for amplifer %s: %f +/- %f", 

1643 amp.getName(), qaStats.getValue(afwMath.MEDIAN), 

1644 qaStats.getValue(afwMath.STDEVCLIP)) 

1645 

1646 self.debugView(ccdExposure, "postISRCCD") 

1647 

1648 return pipeBase.Struct( 

1649 exposure=ccdExposure, 

1650 ossThumb=ossThumb, 

1651 flattenedThumb=flattenedThumb, 

1652 

1653 preInterpolatedExposure=preInterpExp, 

1654 outputExposure=ccdExposure, 

1655 outputOssThumbnail=ossThumb, 

1656 outputFlattenedThumbnail=flattenedThumb, 

1657 ) 

1658 

1659 @pipeBase.timeMethod 

1660 def runDataRef(self, sensorRef): 

1661 """Perform instrument signature removal on a ButlerDataRef of a Sensor. 

1662 

1663 This method contains the `CmdLineTask` interface to the ISR 

1664 processing. All IO is handled here, freeing the `run()` method 

1665 to manage only pixel-level calculations. The steps performed 

1666 are: 

1667 - Read in necessary detrending/isr/calibration data. 

1668 - Process raw exposure in `run()`. 

1669 - Persist the ISR-corrected exposure as "postISRCCD" if 

1670 config.doWrite=True. 

1671 

1672 Parameters 

1673 ---------- 

1674 sensorRef : `daf.persistence.butlerSubset.ButlerDataRef` 

1675 DataRef of the detector data to be processed 

1676 

1677 Returns 

1678 ------- 

1679 result : `lsst.pipe.base.Struct` 

1680 Result struct with component: 

1681 - ``exposure`` : `afw.image.Exposure` 

1682 The fully ISR corrected exposure. 

1683 

1684 Raises 

1685 ------ 

1686 RuntimeError 

1687 Raised if a configuration option is set to True, but the 

1688 required calibration data does not exist. 

1689 

1690 """ 

1691 self.log.info("Performing ISR on sensor %s.", sensorRef.dataId) 

1692 

1693 ccdExposure = sensorRef.get(self.config.datasetType) 

1694 

1695 camera = sensorRef.get("camera") 

1696 isrData = self.readIsrData(sensorRef, ccdExposure) 

1697 

1698 result = self.run(ccdExposure, camera=camera, **isrData.getDict()) 

1699 

1700 if self.config.doWrite: 

1701 sensorRef.put(result.exposure, "postISRCCD") 

1702 if result.preInterpolatedExposure is not None: 

1703 sensorRef.put(result.preInterpolatedExposure, "postISRCCD_uninterpolated") 

1704 if result.ossThumb is not None: 

1705 isrQa.writeThumbnail(sensorRef, result.ossThumb, "ossThumb") 

1706 if result.flattenedThumb is not None: 

1707 isrQa.writeThumbnail(sensorRef, result.flattenedThumb, "flattenedThumb") 

1708 

1709 return result 

1710 

1711 def getIsrExposure(self, dataRef, datasetType, dateObs=None, immediate=True): 

1712 """Retrieve a calibration dataset for removing instrument signature. 

1713 

1714 Parameters 

1715 ---------- 

1716 

1717 dataRef : `daf.persistence.butlerSubset.ButlerDataRef` 

1718 DataRef of the detector data to find calibration datasets 

1719 for. 

1720 datasetType : `str` 

1721 Type of dataset to retrieve (e.g. 'bias', 'flat', etc). 

1722 dateObs : `str`, optional 

1723 Date of the observation. Used to correct butler failures 

1724 when using fallback filters. 

1725 immediate : `Bool` 

1726 If True, disable butler proxies to enable error handling 

1727 within this routine. 

1728 

1729 Returns 

1730 ------- 

1731 exposure : `lsst.afw.image.Exposure` 

1732 Requested calibration frame. 

1733 

1734 Raises 

1735 ------ 

1736 RuntimeError 

1737 Raised if no matching calibration frame can be found. 

1738 """ 

1739 try: 

1740 exp = dataRef.get(datasetType, immediate=immediate) 

1741 except Exception as exc1: 

1742 if not self.config.fallbackFilterName: 

1743 raise RuntimeError("Unable to retrieve %s for %s: %s." % (datasetType, dataRef.dataId, exc1)) 

1744 try: 

1745 if self.config.useFallbackDate and dateObs: 

1746 exp = dataRef.get(datasetType, filter=self.config.fallbackFilterName, 

1747 dateObs=dateObs, immediate=immediate) 

1748 else: 

1749 exp = dataRef.get(datasetType, filter=self.config.fallbackFilterName, immediate=immediate) 

1750 except Exception as exc2: 

1751 raise RuntimeError("Unable to retrieve %s for %s, even with fallback filter %s: %s AND %s." % 

1752 (datasetType, dataRef.dataId, self.config.fallbackFilterName, exc1, exc2)) 

1753 self.log.warn("Using fallback calibration from filter %s.", self.config.fallbackFilterName) 

1754 

1755 if self.config.doAssembleIsrExposures: 

1756 exp = self.assembleCcd.assembleCcd(exp) 

1757 return exp 

1758 

1759 def ensureExposure(self, inputExp, camera, detectorNum): 

1760 """Ensure that the data returned by Butler is a fully constructed exposure. 

1761 

1762 ISR requires exposure-level image data for historical reasons, so if we did 

1763 not recieve that from Butler, construct it from what we have, modifying the 

1764 input in place. 

1765 

1766 Parameters 

1767 ---------- 

1768 inputExp : `lsst.afw.image.Exposure`, `lsst.afw.image.DecoratedImageU`, or 

1769 `lsst.afw.image.ImageF` 

1770 The input data structure obtained from Butler. 

1771 camera : `lsst.afw.cameraGeom.camera` 

1772 The camera associated with the image. Used to find the appropriate 

1773 detector. 

1774 detectorNum : `int` 

1775 The detector this exposure should match. 

1776 

1777 Returns 

1778 ------- 

1779 inputExp : `lsst.afw.image.Exposure` 

1780 The re-constructed exposure, with appropriate detector parameters. 

1781 

1782 Raises 

1783 ------ 

1784 TypeError 

1785 Raised if the input data cannot be used to construct an exposure. 

1786 """ 

1787 if isinstance(inputExp, afwImage.DecoratedImageU): 

1788 inputExp = afwImage.makeExposure(afwImage.makeMaskedImage(inputExp)) 

1789 elif isinstance(inputExp, afwImage.ImageF): 

1790 inputExp = afwImage.makeExposure(afwImage.makeMaskedImage(inputExp)) 

1791 elif isinstance(inputExp, afwImage.MaskedImageF): 

1792 inputExp = afwImage.makeExposure(inputExp) 

1793 elif isinstance(inputExp, afwImage.Exposure): 

1794 pass 

1795 elif inputExp is None: 

1796 # Assume this will be caught by the setup if it is a problem. 

1797 return inputExp 

1798 else: 

1799 raise TypeError("Input Exposure is not known type in isrTask.ensureExposure: %s." % 

1800 (type(inputExp), )) 

1801 

1802 if inputExp.getDetector() is None: 

1803 inputExp.setDetector(camera[detectorNum]) 

1804 

1805 return inputExp 

1806 

1807 def convertIntToFloat(self, exposure): 

1808 """Convert exposure image from uint16 to float. 

1809 

1810 If the exposure does not need to be converted, the input is 

1811 immediately returned. For exposures that are converted to use 

1812 floating point pixels, the variance is set to unity and the 

1813 mask to zero. 

1814 

1815 Parameters 

1816 ---------- 

1817 exposure : `lsst.afw.image.Exposure` 

1818 The raw exposure to be converted. 

1819 

1820 Returns 

1821 ------- 

1822 newexposure : `lsst.afw.image.Exposure` 

1823 The input ``exposure``, converted to floating point pixels. 

1824 

1825 Raises 

1826 ------ 

1827 RuntimeError 

1828 Raised if the exposure type cannot be converted to float. 

1829 

1830 """ 

1831 if isinstance(exposure, afwImage.ExposureF): 

1832 # Nothing to be done 

1833 self.log.debug("Exposure already of type float.") 

1834 return exposure 

1835 if not hasattr(exposure, "convertF"): 

1836 raise RuntimeError("Unable to convert exposure (%s) to float." % type(exposure)) 

1837 

1838 newexposure = exposure.convertF() 

1839 newexposure.variance[:] = 1 

1840 newexposure.mask[:] = 0x0 

1841 

1842 return newexposure 

1843 

1844 def maskAmplifier(self, ccdExposure, amp, defects): 

1845 """Identify bad amplifiers, saturated and suspect pixels. 

1846 

1847 Parameters 

1848 ---------- 

1849 ccdExposure : `lsst.afw.image.Exposure` 

1850 Input exposure to be masked. 

1851 amp : `lsst.afw.table.AmpInfoCatalog` 

1852 Catalog of parameters defining the amplifier on this 

1853 exposure to mask. 

1854 defects : `lsst.ip.isr.Defects` 

1855 List of defects. Used to determine if the entire 

1856 amplifier is bad. 

1857 

1858 Returns 

1859 ------- 

1860 badAmp : `Bool` 

1861 If this is true, the entire amplifier area is covered by 

1862 defects and unusable. 

1863 

1864 """ 

1865 maskedImage = ccdExposure.getMaskedImage() 

1866 

1867 badAmp = False 

1868 

1869 # Check if entire amp region is defined as a defect (need to use amp.getBBox() for correct 

1870 # comparison with current defects definition. 

1871 if defects is not None: 

1872 badAmp = bool(sum([v.getBBox().contains(amp.getBBox()) for v in defects])) 

1873 

1874 # In the case of a bad amp, we will set mask to "BAD" (here use amp.getRawBBox() for correct 

1875 # association with pixels in current ccdExposure). 

1876 if badAmp: 

1877 dataView = afwImage.MaskedImageF(maskedImage, amp.getRawBBox(), 

1878 afwImage.PARENT) 

1879 maskView = dataView.getMask() 

1880 maskView |= maskView.getPlaneBitMask("BAD") 

1881 del maskView 

1882 return badAmp 

1883 

1884 # Mask remaining defects after assembleCcd() to allow for defects that cross amplifier boundaries. 

1885 # Saturation and suspect pixels can be masked now, though. 

1886 limits = dict() 

1887 if self.config.doSaturation and not badAmp: 

1888 limits.update({self.config.saturatedMaskName: amp.getSaturation()}) 

1889 if self.config.doSuspect and not badAmp: 

1890 limits.update({self.config.suspectMaskName: amp.getSuspectLevel()}) 

1891 if math.isfinite(self.config.saturation): 

1892 limits.update({self.config.saturatedMaskName: self.config.saturation}) 

1893 

1894 for maskName, maskThreshold in limits.items(): 

1895 if not math.isnan(maskThreshold): 

1896 dataView = maskedImage.Factory(maskedImage, amp.getRawBBox()) 

1897 isrFunctions.makeThresholdMask( 

1898 maskedImage=dataView, 

1899 threshold=maskThreshold, 

1900 growFootprints=0, 

1901 maskName=maskName 

1902 ) 

1903 

1904 # Determine if we've fully masked this amplifier with SUSPECT and SAT pixels. 

1905 maskView = afwImage.Mask(maskedImage.getMask(), amp.getRawDataBBox(), 

1906 afwImage.PARENT) 

1907 maskVal = maskView.getPlaneBitMask([self.config.saturatedMaskName, 

1908 self.config.suspectMaskName]) 

1909 if numpy.all(maskView.getArray() & maskVal > 0): 

1910 badAmp = True 

1911 maskView |= maskView.getPlaneBitMask("BAD") 

1912 

1913 return badAmp 

1914 

1915 def overscanCorrection(self, ccdExposure, amp): 

1916 """Apply overscan correction in place. 

1917 

1918 This method does initial pixel rejection of the overscan 

1919 region. The overscan can also be optionally segmented to 

1920 allow for discontinuous overscan responses to be fit 

1921 separately. The actual overscan subtraction is performed by 

1922 the `lsst.ip.isr.isrFunctions.overscanCorrection` function, 

1923 which is called here after the amplifier is preprocessed. 

1924 

1925 Parameters 

1926 ---------- 

1927 ccdExposure : `lsst.afw.image.Exposure` 

1928 Exposure to have overscan correction performed. 

1929 amp : `lsst.afw.cameraGeom.Amplifer` 

1930 The amplifier to consider while correcting the overscan. 

1931 

1932 Returns 

1933 ------- 

1934 overscanResults : `lsst.pipe.base.Struct` 

1935 Result struct with components: 

1936 - ``imageFit`` : scalar or `lsst.afw.image.Image` 

1937 Value or fit subtracted from the amplifier image data. 

1938 - ``overscanFit`` : scalar or `lsst.afw.image.Image` 

1939 Value or fit subtracted from the overscan image data. 

1940 - ``overscanImage`` : `lsst.afw.image.Image` 

1941 Image of the overscan region with the overscan 

1942 correction applied. This quantity is used to estimate 

1943 the amplifier read noise empirically. 

1944 

1945 Raises 

1946 ------ 

1947 RuntimeError 

1948 Raised if the ``amp`` does not contain raw pixel information. 

1949 

1950 See Also 

1951 -------- 

1952 lsst.ip.isr.isrFunctions.overscanCorrection 

1953 """ 

1954 if amp.getRawHorizontalOverscanBBox().isEmpty(): 

1955 self.log.info("ISR_OSCAN: No overscan region. Not performing overscan correction.") 

1956 return None 

1957 

1958 statControl = afwMath.StatisticsControl() 

1959 statControl.setAndMask(ccdExposure.mask.getPlaneBitMask("SAT")) 

1960 

1961 # Determine the bounding boxes 

1962 dataBBox = amp.getRawDataBBox() 

1963 oscanBBox = amp.getRawHorizontalOverscanBBox() 

1964 dx0 = 0 

1965 dx1 = 0 

1966 

1967 prescanBBox = amp.getRawPrescanBBox() 

1968 if (oscanBBox.getBeginX() > prescanBBox.getBeginX()): # amp is at the right 

1969 dx0 += self.config.overscanNumLeadingColumnsToSkip 

1970 dx1 -= self.config.overscanNumTrailingColumnsToSkip 

1971 else: 

1972 dx0 += self.config.overscanNumTrailingColumnsToSkip 

1973 dx1 -= self.config.overscanNumLeadingColumnsToSkip 

1974 

1975 # Determine if we need to work on subregions of the amplifier and overscan. 

1976 imageBBoxes = [] 

1977 overscanBBoxes = [] 

1978 

1979 if ((self.config.overscanBiasJump 

1980 and self.config.overscanBiasJumpLocation) 

1981 and (ccdExposure.getMetadata().exists(self.config.overscanBiasJumpKeyword) 

1982 and ccdExposure.getMetadata().getScalar(self.config.overscanBiasJumpKeyword) in 

1983 self.config.overscanBiasJumpDevices)): 

1984 if amp.getReadoutCorner() in (ReadoutCorner.LL, ReadoutCorner.LR): 

1985 yLower = self.config.overscanBiasJumpLocation 

1986 yUpper = dataBBox.getHeight() - yLower 

1987 else: 

1988 yUpper = self.config.overscanBiasJumpLocation 

1989 yLower = dataBBox.getHeight() - yUpper 

1990 

1991 imageBBoxes.append(lsst.geom.Box2I(dataBBox.getBegin(), 

1992 lsst.geom.Extent2I(dataBBox.getWidth(), yLower))) 

1993 overscanBBoxes.append(lsst.geom.Box2I(oscanBBox.getBegin() + lsst.geom.Extent2I(dx0, 0), 

1994 lsst.geom.Extent2I(oscanBBox.getWidth() - dx0 + dx1, 

1995 yLower))) 

1996 

1997 imageBBoxes.append(lsst.geom.Box2I(dataBBox.getBegin() + lsst.geom.Extent2I(0, yLower), 

1998 lsst.geom.Extent2I(dataBBox.getWidth(), yUpper))) 

1999 overscanBBoxes.append(lsst.geom.Box2I(oscanBBox.getBegin() + lsst.geom.Extent2I(dx0, yLower), 

2000 lsst.geom.Extent2I(oscanBBox.getWidth() - dx0 + dx1, 

2001 yUpper))) 

2002 else: 

2003 imageBBoxes.append(lsst.geom.Box2I(dataBBox.getBegin(), 

2004 lsst.geom.Extent2I(dataBBox.getWidth(), dataBBox.getHeight()))) 

2005 overscanBBoxes.append(lsst.geom.Box2I(oscanBBox.getBegin() + lsst.geom.Extent2I(dx0, 0), 

2006 lsst.geom.Extent2I(oscanBBox.getWidth() - dx0 + dx1, 

2007 oscanBBox.getHeight()))) 

2008 

2009 # Perform overscan correction on subregions, ensuring saturated pixels are masked. 

2010 for imageBBox, overscanBBox in zip(imageBBoxes, overscanBBoxes): 

2011 ampImage = ccdExposure.maskedImage[imageBBox] 

2012 overscanImage = ccdExposure.maskedImage[overscanBBox] 

2013 

2014 overscanArray = overscanImage.image.array 

2015 median = numpy.ma.median(numpy.ma.masked_where(overscanImage.mask.array, overscanArray)) 

2016 bad = numpy.where(numpy.abs(overscanArray - median) > self.config.overscanMaxDev) 

2017 overscanImage.mask.array[bad] = overscanImage.mask.getPlaneBitMask("SAT") 

2018 

2019 statControl = afwMath.StatisticsControl() 

2020 statControl.setAndMask(ccdExposure.mask.getPlaneBitMask("SAT")) 

2021 

2022 overscanResults = self.overscan.run(ampImage.getImage(), overscanImage, amp) 

2023 

2024 # Measure average overscan levels and record them in the metadata. 

2025 levelStat = afwMath.MEDIAN 

2026 sigmaStat = afwMath.STDEVCLIP 

2027 

2028 sctrl = afwMath.StatisticsControl(self.config.qa.flatness.clipSigma, 

2029 self.config.qa.flatness.nIter) 

2030 metadata = ccdExposure.getMetadata() 

2031 ampNum = amp.getName() 

2032 # if self.config.overscanFitType in ("MEDIAN", "MEAN", "MEANCLIP"): 

2033 if isinstance(overscanResults.overscanFit, float): 

2034 metadata.set("ISR_OSCAN_LEVEL%s" % ampNum, overscanResults.overscanFit) 

2035 metadata.set("ISR_OSCAN_SIGMA%s" % ampNum, 0.0) 

2036 else: 

2037 stats = afwMath.makeStatistics(overscanResults.overscanFit, levelStat | sigmaStat, sctrl) 

2038 metadata.set("ISR_OSCAN_LEVEL%s" % ampNum, stats.getValue(levelStat)) 

2039 metadata.set("ISR_OSCAN_SIGMA%s" % ampNum, stats.getValue(sigmaStat)) 

2040 

2041 return overscanResults 

2042 

2043 def updateVariance(self, ampExposure, amp, overscanImage=None): 

2044 """Set the variance plane using the amplifier gain and read noise 

2045 

2046 The read noise is calculated from the ``overscanImage`` if the 

2047 ``doEmpiricalReadNoise`` option is set in the configuration; otherwise 

2048 the value from the amplifier data is used. 

2049 

2050 Parameters 

2051 ---------- 

2052 ampExposure : `lsst.afw.image.Exposure` 

2053 Exposure to process. 

2054 amp : `lsst.afw.table.AmpInfoRecord` or `FakeAmp` 

2055 Amplifier detector data. 

2056 overscanImage : `lsst.afw.image.MaskedImage`, optional. 

2057 Image of overscan, required only for empirical read noise. 

2058 

2059 See also 

2060 -------- 

2061 lsst.ip.isr.isrFunctions.updateVariance 

2062 """ 

2063 maskPlanes = [self.config.saturatedMaskName, self.config.suspectMaskName] 

2064 gain = amp.getGain() 

2065 

2066 if math.isnan(gain): 

2067 gain = 1.0 

2068 self.log.warn("Gain set to NAN! Updating to 1.0 to generate Poisson variance.") 

2069 elif gain <= 0: 

2070 patchedGain = 1.0 

2071 self.log.warn("Gain for amp %s == %g <= 0; setting to %f.", 

2072 amp.getName(), gain, patchedGain) 

2073 gain = patchedGain 

2074 

2075 if self.config.doEmpiricalReadNoise and overscanImage is None: 

2076 self.log.info("Overscan is none for EmpiricalReadNoise.") 

2077 

2078 if self.config.doEmpiricalReadNoise and overscanImage is not None: 

2079 stats = afwMath.StatisticsControl() 

2080 stats.setAndMask(overscanImage.mask.getPlaneBitMask(maskPlanes)) 

2081 readNoise = afwMath.makeStatistics(overscanImage, afwMath.STDEVCLIP, stats).getValue() 

2082 self.log.info("Calculated empirical read noise for amp %s: %f.", 

2083 amp.getName(), readNoise) 

2084 else: 

2085 readNoise = amp.getReadNoise() 

2086 

2087 isrFunctions.updateVariance( 

2088 maskedImage=ampExposure.getMaskedImage(), 

2089 gain=gain, 

2090 readNoise=readNoise, 

2091 ) 

2092 

2093 def darkCorrection(self, exposure, darkExposure, invert=False): 

2094 """Apply dark correction in place. 

2095 

2096 Parameters 

2097 ---------- 

2098 exposure : `lsst.afw.image.Exposure` 

2099 Exposure to process. 

2100 darkExposure : `lsst.afw.image.Exposure` 

2101 Dark exposure of the same size as ``exposure``. 

2102 invert : `Bool`, optional 

2103 If True, re-add the dark to an already corrected image. 

2104 

2105 Raises 

2106 ------ 

2107 RuntimeError 

2108 Raised if either ``exposure`` or ``darkExposure`` do not 

2109 have their dark time defined. 

2110 

2111 See Also 

2112 -------- 

2113 lsst.ip.isr.isrFunctions.darkCorrection 

2114 """ 

2115 expScale = exposure.getInfo().getVisitInfo().getDarkTime() 

2116 if math.isnan(expScale): 

2117 raise RuntimeError("Exposure darktime is NAN.") 

2118 if darkExposure.getInfo().getVisitInfo() is not None \ 

2119 and not math.isnan(darkExposure.getInfo().getVisitInfo().getDarkTime()): 

2120 darkScale = darkExposure.getInfo().getVisitInfo().getDarkTime() 

2121 else: 

2122 # DM-17444: darkExposure.getInfo.getVisitInfo() is None 

2123 # so getDarkTime() does not exist. 

2124 self.log.warn("darkExposure.getInfo().getVisitInfo() does not exist. Using darkScale = 1.0.") 

2125 darkScale = 1.0 

2126 

2127 isrFunctions.darkCorrection( 

2128 maskedImage=exposure.getMaskedImage(), 

2129 darkMaskedImage=darkExposure.getMaskedImage(), 

2130 expScale=expScale, 

2131 darkScale=darkScale, 

2132 invert=invert, 

2133 trimToFit=self.config.doTrimToMatchCalib 

2134 ) 

2135 

2136 def doLinearize(self, detector): 

2137 """Check if linearization is needed for the detector cameraGeom. 

2138 

2139 Checks config.doLinearize and the linearity type of the first 

2140 amplifier. 

2141 

2142 Parameters 

2143 ---------- 

2144 detector : `lsst.afw.cameraGeom.Detector` 

2145 Detector to get linearity type from. 

2146 

2147 Returns 

2148 ------- 

2149 doLinearize : `Bool` 

2150 If True, linearization should be performed. 

2151 """ 

2152 return self.config.doLinearize and \ 

2153 detector.getAmplifiers()[0].getLinearityType() != NullLinearityType 

2154 

2155 def flatCorrection(self, exposure, flatExposure, invert=False): 

2156 """Apply flat correction in place. 

2157 

2158 Parameters 

2159 ---------- 

2160 exposure : `lsst.afw.image.Exposure` 

2161 Exposure to process. 

2162 flatExposure : `lsst.afw.image.Exposure` 

2163 Flat exposure of the same size as ``exposure``. 

2164 invert : `Bool`, optional 

2165 If True, unflatten an already flattened image. 

2166 

2167 See Also 

2168 -------- 

2169 lsst.ip.isr.isrFunctions.flatCorrection 

2170 """ 

2171 isrFunctions.flatCorrection( 

2172 maskedImage=exposure.getMaskedImage(), 

2173 flatMaskedImage=flatExposure.getMaskedImage(), 

2174 scalingType=self.config.flatScalingType, 

2175 userScale=self.config.flatUserScale, 

2176 invert=invert, 

2177 trimToFit=self.config.doTrimToMatchCalib 

2178 ) 

2179 

2180 def saturationDetection(self, exposure, amp): 

2181 """Detect saturated pixels and mask them using mask plane config.saturatedMaskName, in place. 

2182 

2183 Parameters 

2184 ---------- 

2185 exposure : `lsst.afw.image.Exposure` 

2186 Exposure to process. Only the amplifier DataSec is processed. 

2187 amp : `lsst.afw.table.AmpInfoCatalog` 

2188 Amplifier detector data. 

2189 

2190 See Also 

2191 -------- 

2192 lsst.ip.isr.isrFunctions.makeThresholdMask 

2193 """ 

2194 if not math.isnan(amp.getSaturation()): 

2195 maskedImage = exposure.getMaskedImage() 

2196 dataView = maskedImage.Factory(maskedImage, amp.getRawBBox()) 

2197 isrFunctions.makeThresholdMask( 

2198 maskedImage=dataView, 

2199 threshold=amp.getSaturation(), 

2200 growFootprints=0, 

2201 maskName=self.config.saturatedMaskName, 

2202 ) 

2203 

2204 def saturationInterpolation(self, exposure): 

2205 """Interpolate over saturated pixels, in place. 

2206 

2207 This method should be called after `saturationDetection`, to 

2208 ensure that the saturated pixels have been identified in the 

2209 SAT mask. It should also be called after `assembleCcd`, since 

2210 saturated regions may cross amplifier boundaries. 

2211 

2212 Parameters 

2213 ---------- 

2214 exposure : `lsst.afw.image.Exposure` 

2215 Exposure to process. 

2216 

2217 See Also 

2218 -------- 

2219 lsst.ip.isr.isrTask.saturationDetection 

2220 lsst.ip.isr.isrFunctions.interpolateFromMask 

2221 """ 

2222 isrFunctions.interpolateFromMask( 

2223 maskedImage=exposure.getMaskedImage(), 

2224 fwhm=self.config.fwhm, 

2225 growSaturatedFootprints=self.config.growSaturationFootprintSize, 

2226 maskNameList=list(self.config.saturatedMaskName), 

2227 ) 

2228 

2229 def suspectDetection(self, exposure, amp): 

2230 """Detect suspect pixels and mask them using mask plane config.suspectMaskName, in place. 

2231 

2232 Parameters 

2233 ---------- 

2234 exposure : `lsst.afw.image.Exposure` 

2235 Exposure to process. Only the amplifier DataSec is processed. 

2236 amp : `lsst.afw.table.AmpInfoCatalog` 

2237 Amplifier detector data. 

2238 

2239 See Also 

2240 -------- 

2241 lsst.ip.isr.isrFunctions.makeThresholdMask 

2242 

2243 Notes 

2244 ----- 

2245 Suspect pixels are pixels whose value is greater than amp.getSuspectLevel(). 

2246 This is intended to indicate pixels that may be affected by unknown systematics; 

2247 for example if non-linearity corrections above a certain level are unstable 

2248 then that would be a useful value for suspectLevel. A value of `nan` indicates 

2249 that no such level exists and no pixels are to be masked as suspicious. 

2250 """ 

2251 suspectLevel = amp.getSuspectLevel() 

2252 if math.isnan(suspectLevel): 

2253 return 

2254 

2255 maskedImage = exposure.getMaskedImage() 

2256 dataView = maskedImage.Factory(maskedImage, amp.getRawBBox()) 

2257 isrFunctions.makeThresholdMask( 

2258 maskedImage=dataView, 

2259 threshold=suspectLevel, 

2260 growFootprints=0, 

2261 maskName=self.config.suspectMaskName, 

2262 ) 

2263 

2264 def maskDefect(self, exposure, defectBaseList): 

2265 """Mask defects using mask plane "BAD", in place. 

2266 

2267 Parameters 

2268 ---------- 

2269 exposure : `lsst.afw.image.Exposure` 

2270 Exposure to process. 

2271 defectBaseList : `lsst.ip.isr.Defects` or `list` of 

2272 `lsst.afw.image.DefectBase`. 

2273 List of defects to mask. 

2274 

2275 Notes 

2276 ----- 

2277 Call this after CCD assembly, since defects may cross amplifier boundaries. 

2278 """ 

2279 maskedImage = exposure.getMaskedImage() 

2280 if not isinstance(defectBaseList, Defects): 

2281 # Promotes DefectBase to Defect 

2282 defectList = Defects(defectBaseList) 

2283 else: 

2284 defectList = defectBaseList 

2285 defectList.maskPixels(maskedImage, maskName="BAD") 

2286 

2287 def maskEdges(self, exposure, numEdgePixels=0, maskPlane="SUSPECT", level='DETECTOR'): 

2288 """Mask edge pixels with applicable mask plane. 

2289 

2290 Parameters 

2291 ---------- 

2292 exposure : `lsst.afw.image.Exposure` 

2293 Exposure to process. 

2294 numEdgePixels : `int`, optional 

2295 Number of edge pixels to mask. 

2296 maskPlane : `str`, optional 

2297 Mask plane name to use. 

2298 level : `str`, optional 

2299 Level at which to mask edges. 

2300 """ 

2301 maskedImage = exposure.getMaskedImage() 

2302 maskBitMask = maskedImage.getMask().getPlaneBitMask(maskPlane) 

2303 

2304 if numEdgePixels > 0: 

2305 if level == 'DETECTOR': 

2306 boxes = [maskedImage.getBBox()] 

2307 elif level == 'AMP': 

2308 boxes = [amp.getBBox() for amp in exposure.getDetector()] 

2309 

2310 for box in boxes: 

2311 # This makes a bbox numEdgeSuspect pixels smaller than the image on each side 

2312 subImage = maskedImage[box] 

2313 box.grow(-numEdgePixels) 

2314 # Mask pixels outside box 

2315 SourceDetectionTask.setEdgeBits( 

2316 subImage, 

2317 box, 

2318 maskBitMask) 

2319 

2320 def maskAndInterpolateDefects(self, exposure, defectBaseList): 

2321 """Mask and interpolate defects using mask plane "BAD", in place. 

2322 

2323 Parameters 

2324 ---------- 

2325 exposure : `lsst.afw.image.Exposure` 

2326 Exposure to process. 

2327 defectBaseList : `lsst.ip.isr.Defects` or `list` of 

2328 `lsst.afw.image.DefectBase`. 

2329 List of defects to mask and interpolate. 

2330 

2331 See Also 

2332 -------- 

2333 lsst.ip.isr.isrTask.maskDefect 

2334 """ 

2335 self.maskDefect(exposure, defectBaseList) 

2336 self.maskEdges(exposure, numEdgePixels=self.config.numEdgeSuspect, 

2337 maskPlane="SUSPECT", level=self.config.edgeMaskLevel) 

2338 isrFunctions.interpolateFromMask( 

2339 maskedImage=exposure.getMaskedImage(), 

2340 fwhm=self.config.fwhm, 

2341 growSaturatedFootprints=0, 

2342 maskNameList=["BAD"], 

2343 ) 

2344 

2345 def maskNan(self, exposure): 

2346 """Mask NaNs using mask plane "UNMASKEDNAN", in place. 

2347 

2348 Parameters 

2349 ---------- 

2350 exposure : `lsst.afw.image.Exposure` 

2351 Exposure to process. 

2352 

2353 Notes 

2354 ----- 

2355 We mask over all NaNs, including those that are masked with 

2356 other bits (because those may or may not be interpolated over 

2357 later, and we want to remove all NaNs). Despite this 

2358 behaviour, the "UNMASKEDNAN" mask plane is used to preserve 

2359 the historical name. 

2360 """ 

2361 maskedImage = exposure.getMaskedImage() 

2362 

2363 # Find and mask NaNs 

2364 maskedImage.getMask().addMaskPlane("UNMASKEDNAN") 

2365 maskVal = maskedImage.getMask().getPlaneBitMask("UNMASKEDNAN") 

2366 numNans = maskNans(maskedImage, maskVal) 

2367 self.metadata.set("NUMNANS", numNans) 

2368 if numNans > 0: 

2369 self.log.warn("There were %d unmasked NaNs.", numNans) 

2370 

2371 def maskAndInterpolateNan(self, exposure): 

2372 """"Mask and interpolate NaNs using mask plane "UNMASKEDNAN", in place. 

2373 

2374 Parameters 

2375 ---------- 

2376 exposure : `lsst.afw.image.Exposure` 

2377 Exposure to process. 

2378 

2379 See Also 

2380 -------- 

2381 lsst.ip.isr.isrTask.maskNan 

2382 """ 

2383 self.maskNan(exposure) 

2384 isrFunctions.interpolateFromMask( 

2385 maskedImage=exposure.getMaskedImage(), 

2386 fwhm=self.config.fwhm, 

2387 growSaturatedFootprints=0, 

2388 maskNameList=["UNMASKEDNAN"], 

2389 ) 

2390 

2391 def measureBackground(self, exposure, IsrQaConfig=None): 

2392 """Measure the image background in subgrids, for quality control purposes. 

2393 

2394 Parameters 

2395 ---------- 

2396 exposure : `lsst.afw.image.Exposure` 

2397 Exposure to process. 

2398 IsrQaConfig : `lsst.ip.isr.isrQa.IsrQaConfig` 

2399 Configuration object containing parameters on which background 

2400 statistics and subgrids to use. 

2401 """ 

2402 if IsrQaConfig is not None: 

2403 statsControl = afwMath.StatisticsControl(IsrQaConfig.flatness.clipSigma, 

2404 IsrQaConfig.flatness.nIter) 

2405 maskVal = exposure.getMaskedImage().getMask().getPlaneBitMask(["BAD", "SAT", "DETECTED"]) 

2406 statsControl.setAndMask(maskVal) 

2407 maskedImage = exposure.getMaskedImage() 

2408 stats = afwMath.makeStatistics(maskedImage, afwMath.MEDIAN | afwMath.STDEVCLIP, statsControl) 

2409 skyLevel = stats.getValue(afwMath.MEDIAN) 

2410 skySigma = stats.getValue(afwMath.STDEVCLIP) 

2411 self.log.info("Flattened sky level: %f +/- %f.", skyLevel, skySigma) 

2412 metadata = exposure.getMetadata() 

2413 metadata.set('SKYLEVEL', skyLevel) 

2414 metadata.set('SKYSIGMA', skySigma) 

2415 

2416 # calcluating flatlevel over the subgrids 

2417 stat = afwMath.MEANCLIP if IsrQaConfig.flatness.doClip else afwMath.MEAN 

2418 meshXHalf = int(IsrQaConfig.flatness.meshX/2.) 

2419 meshYHalf = int(IsrQaConfig.flatness.meshY/2.) 

2420 nX = int((exposure.getWidth() + meshXHalf) / IsrQaConfig.flatness.meshX) 

2421 nY = int((exposure.getHeight() + meshYHalf) / IsrQaConfig.flatness.meshY) 

2422 skyLevels = numpy.zeros((nX, nY)) 

2423 

2424 for j in range(nY): 

2425 yc = meshYHalf + j * IsrQaConfig.flatness.meshY 

2426 for i in range(nX): 

2427 xc = meshXHalf + i * IsrQaConfig.flatness.meshX 

2428 

2429 xLLC = xc - meshXHalf 

2430 yLLC = yc - meshYHalf 

2431 xURC = xc + meshXHalf - 1 

2432 yURC = yc + meshYHalf - 1 

2433 

2434 bbox = lsst.geom.Box2I(lsst.geom.Point2I(xLLC, yLLC), lsst.geom.Point2I(xURC, yURC)) 

2435 miMesh = maskedImage.Factory(exposure.getMaskedImage(), bbox, afwImage.LOCAL) 

2436 

2437 skyLevels[i, j] = afwMath.makeStatistics(miMesh, stat, statsControl).getValue() 

2438 

2439 good = numpy.where(numpy.isfinite(skyLevels)) 

2440 skyMedian = numpy.median(skyLevels[good]) 

2441 flatness = (skyLevels[good] - skyMedian) / skyMedian 

2442 flatness_rms = numpy.std(flatness) 

2443 flatness_pp = flatness.max() - flatness.min() if len(flatness) > 0 else numpy.nan 

2444 

2445 self.log.info("Measuring sky levels in %dx%d grids: %f.", nX, nY, skyMedian) 

2446 self.log.info("Sky flatness in %dx%d grids - pp: %f rms: %f.", 

2447 nX, nY, flatness_pp, flatness_rms) 

2448 

2449 metadata.set('FLATNESS_PP', float(flatness_pp)) 

2450 metadata.set('FLATNESS_RMS', float(flatness_rms)) 

2451 metadata.set('FLATNESS_NGRIDS', '%dx%d' % (nX, nY)) 

2452 metadata.set('FLATNESS_MESHX', IsrQaConfig.flatness.meshX) 

2453 metadata.set('FLATNESS_MESHY', IsrQaConfig.flatness.meshY) 

2454 

2455 def roughZeroPoint(self, exposure): 

2456 """Set an approximate magnitude zero point for the exposure. 

2457 

2458 Parameters 

2459 ---------- 

2460 exposure : `lsst.afw.image.Exposure` 

2461 Exposure to process. 

2462 """ 

2463 filterName = afwImage.Filter(exposure.getFilter().getId()).getName() # Canonical name for filter 

2464 if filterName in self.config.fluxMag0T1: 

2465 fluxMag0 = self.config.fluxMag0T1[filterName] 

2466 else: 

2467 self.log.warn("No rough magnitude zero point set for filter %s.", filterName) 

2468 fluxMag0 = self.config.defaultFluxMag0T1 

2469 

2470 expTime = exposure.getInfo().getVisitInfo().getExposureTime() 

2471 if not expTime > 0: # handle NaN as well as <= 0 

2472 self.log.warn("Non-positive exposure time; skipping rough zero point.") 

2473 return 

2474 

2475 self.log.info("Setting rough magnitude zero point: %f", 2.5*math.log10(fluxMag0*expTime)) 

2476 exposure.setPhotoCalib(afwImage.makePhotoCalibFromCalibZeroPoint(fluxMag0*expTime, 0.0)) 

2477 

2478 def setValidPolygonIntersect(self, ccdExposure, fpPolygon): 

2479 """Set the valid polygon as the intersection of fpPolygon and the ccd corners. 

2480 

2481 Parameters 

2482 ---------- 

2483 ccdExposure : `lsst.afw.image.Exposure` 

2484 Exposure to process. 

2485 fpPolygon : `lsst.afw.geom.Polygon` 

2486 Polygon in focal plane coordinates. 

2487 """ 

2488 # Get ccd corners in focal plane coordinates 

2489 ccd = ccdExposure.getDetector() 

2490 fpCorners = ccd.getCorners(FOCAL_PLANE) 

2491 ccdPolygon = Polygon(fpCorners) 

2492 

2493 # Get intersection of ccd corners with fpPolygon 

2494 intersect = ccdPolygon.intersectionSingle(fpPolygon) 

2495 

2496 # Transform back to pixel positions and build new polygon 

2497 ccdPoints = ccd.transform(intersect, FOCAL_PLANE, PIXELS) 

2498 validPolygon = Polygon(ccdPoints) 

2499 ccdExposure.getInfo().setValidPolygon(validPolygon) 

2500 

2501 @contextmanager 

2502 def flatContext(self, exp, flat, dark=None): 

2503 """Context manager that applies and removes flats and darks, 

2504 if the task is configured to apply them. 

2505 

2506 Parameters 

2507 ---------- 

2508 exp : `lsst.afw.image.Exposure` 

2509 Exposure to process. 

2510 flat : `lsst.afw.image.Exposure` 

2511 Flat exposure the same size as ``exp``. 

2512 dark : `lsst.afw.image.Exposure`, optional 

2513 Dark exposure the same size as ``exp``. 

2514 

2515 Yields 

2516 ------ 

2517 exp : `lsst.afw.image.Exposure` 

2518 The flat and dark corrected exposure. 

2519 """ 

2520 if self.config.doDark and dark is not None: 

2521 self.darkCorrection(exp, dark) 

2522 if self.config.doFlat: 

2523 self.flatCorrection(exp, flat) 

2524 try: 

2525 yield exp 

2526 finally: 

2527 if self.config.doFlat: 

2528 self.flatCorrection(exp, flat, invert=True) 

2529 if self.config.doDark and dark is not None: 

2530 self.darkCorrection(exp, dark, invert=True) 

2531 

2532 def debugView(self, exposure, stepname): 

2533 """Utility function to examine ISR exposure at different stages. 

2534 

2535 Parameters 

2536 ---------- 

2537 exposure : `lsst.afw.image.Exposure` 

2538 Exposure to view. 

2539 stepname : `str` 

2540 State of processing to view. 

2541 """ 

2542 frame = getDebugFrame(self._display, stepname) 

2543 if frame: 

2544 display = getDisplay(frame) 

2545 display.scale('asinh', 'zscale') 

2546 display.mtv(exposure) 

2547 prompt = "Press Enter to continue [c]... " 

2548 while True: 

2549 ans = input(prompt).lower() 

2550 if ans in ("", "c",): 

2551 break 

2552 

2553 

2554class FakeAmp(object): 

2555 """A Detector-like object that supports returning gain and saturation level 

2556 

2557 This is used when the input exposure does not have a detector. 

2558 

2559 Parameters 

2560 ---------- 

2561 exposure : `lsst.afw.image.Exposure` 

2562 Exposure to generate a fake amplifier for. 

2563 config : `lsst.ip.isr.isrTaskConfig` 

2564 Configuration to apply to the fake amplifier. 

2565 """ 

2566 

2567 def __init__(self, exposure, config): 

2568 self._bbox = exposure.getBBox(afwImage.LOCAL) 

2569 self._RawHorizontalOverscanBBox = lsst.geom.Box2I() 

2570 self._gain = config.gain 

2571 self._readNoise = config.readNoise 

2572 self._saturation = config.saturation 

2573 

2574 def getBBox(self): 

2575 return self._bbox 

2576 

2577 def getRawBBox(self): 

2578 return self._bbox 

2579 

2580 def getRawHorizontalOverscanBBox(self): 

2581 return self._RawHorizontalOverscanBBox 

2582 

2583 def getGain(self): 

2584 return self._gain 

2585 

2586 def getReadNoise(self): 

2587 return self._readNoise 

2588 

2589 def getSaturation(self): 

2590 return self._saturation 

2591 

2592 def getSuspectLevel(self): 

2593 return float("NaN") 

2594 

2595 

2596class RunIsrConfig(pexConfig.Config): 

2597 isr = pexConfig.ConfigurableField(target=IsrTask, doc="Instrument signature removal") 

2598 

2599 

2600class RunIsrTask(pipeBase.CmdLineTask): 

2601 """Task to wrap the default IsrTask to allow it to be retargeted. 

2602 

2603 The standard IsrTask can be called directly from a command line 

2604 program, but doing so removes the ability of the task to be 

2605 retargeted. As most cameras override some set of the IsrTask 

2606 methods, this would remove those data-specific methods in the 

2607 output post-ISR images. This wrapping class fixes the issue, 

2608 allowing identical post-ISR images to be generated by both the 

2609 processCcd and isrTask code. 

2610 """ 

2611 ConfigClass = RunIsrConfig 

2612 _DefaultName = "runIsr" 

2613 

2614 def __init__(self, *args, **kwargs): 

2615 super().__init__(*args, **kwargs) 

2616 self.makeSubtask("isr") 

2617 

2618 def runDataRef(self, dataRef): 

2619 """ 

2620 Parameters 

2621 ---------- 

2622 dataRef : `lsst.daf.persistence.ButlerDataRef` 

2623 data reference of the detector data to be processed 

2624 

2625 Returns 

2626 ------- 

2627 result : `pipeBase.Struct` 

2628 Result struct with component: 

2629 

2630 - exposure : `lsst.afw.image.Exposure` 

2631 Post-ISR processed exposure. 

2632 """ 

2633 return self.isr.runDataRef(dataRef)