lsst.ip.diffim gf0d7cb17a7+bca286fd3d
Loading...
Searching...
No Matches
dipoleFitTask.py
Go to the documentation of this file.
2# LSST Data Management System
3# Copyright 2008-2016 AURA/LSST.
4#
5# This product includes software developed by the
6# LSST Project (http://www.lsst.org/).
7#
8# This program is free software: you can redistribute it and/or modify
9# it under the terms of the GNU General Public License as published by
10# the Free Software Foundation, either version 3 of the License, or
11# (at your option) any later version.
12#
13# This program is distributed in the hope that it will be useful,
14# but WITHOUT ANY WARRANTY; without even the implied warranty of
15# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16# GNU General Public License for more details.
17#
18# You should have received a copy of the LSST License Statement and
19# the GNU General Public License along with this program. If not,
20# see <https://www.lsstcorp.org/LegalNotices/>.
21#
22
23import logging
24import numpy as np
25import warnings
26
27import lsst.afw.image as afwImage
28import lsst.meas.base as measBase
29import lsst.afw.table as afwTable
30import lsst.afw.detection as afwDet
31import lsst.geom as geom
32import lsst.pex.exceptions as pexExcept
33import lsst.pex.config as pexConfig
34from lsst.pipe.base import Struct
35from lsst.utils.timer import timeMethod
36
37__all__ = ("DipoleFitTask", "DipoleFitPlugin", "DipoleFitTaskConfig", "DipoleFitPluginConfig",
38 "DipoleFitAlgorithm")
39
40
41# Create a new measurement task (`DipoleFitTask`) that can handle all other SFM tasks but can
42# pass a separate pos- and neg- exposure/image to the `DipoleFitPlugin`s `run()` method.
43
44
45class DipoleFitPluginConfig(measBase.SingleFramePluginConfig):
46 """Configuration for DipoleFitPlugin
47 """
48
49 fitAllDiaSources = pexConfig.Field(
50 dtype=float, default=False,
51 doc="""Attempte dipole fit of all diaSources (otherwise just the ones consisting of overlapping
52 positive and negative footprints)""")
53
54 maxSeparation = pexConfig.Field(
55 dtype=float, default=5.,
56 doc="Assume dipole is not separated by more than maxSeparation * psfSigma")
57
58 relWeight = pexConfig.Field(
59 dtype=float, default=0.5,
60 doc="""Relative weighting of pre-subtraction images (higher -> greater influence of pre-sub.
61 images on fit)""")
62
63 tolerance = pexConfig.Field(
64 dtype=float, default=1e-7,
65 doc="Fit tolerance")
66
67 fitBackground = pexConfig.Field(
68 dtype=int, default=1,
69 doc="Set whether and how to fit for linear gradient in pre-sub. images. Possible values:"
70 "0: do not fit background at all"
71 "1 (default): pre-fit the background using linear least squares and then do not fit it as part"
72 "of the dipole fitting optimization"
73 "2: pre-fit the background using linear least squares (as in 1), and use the parameter"
74 "estimates from that fit as starting parameters for an integrated re-fit of the background")
75
76 fitSeparateNegParams = pexConfig.Field(
77 dtype=bool, default=False,
78 doc="Include parameters to fit for negative values (flux, gradient) separately from pos.")
79
80 # Config params for classification of detected diaSources as dipole or not
81 minSn = pexConfig.Field(
82 dtype=float, default=np.sqrt(2) * 5.0,
83 doc="Minimum quadrature sum of positive+negative lobe S/N to be considered a dipole")
84
85 maxFluxRatio = pexConfig.Field(
86 dtype=float, default=0.65,
87 doc="Maximum flux ratio in either lobe to be considered a dipole")
88
89 maxChi2DoF = pexConfig.Field(
90 dtype=float, default=0.05,
91 doc="""Maximum Chi2/DoF significance of fit to be considered a dipole.
92 Default value means \"Choose a chi2DoF corresponding to a significance level of at most 0.05\"
93 (note this is actually a significance, not a chi2 value).""")
94
95
96class DipoleFitTaskConfig(measBase.SingleFrameMeasurementConfig):
97 """Measurement of detected diaSources as dipoles
98
99 Currently we keep the "old" DipoleMeasurement algorithms turned on.
100 """
101
102 def setDefaults(self):
103 measBase.SingleFrameMeasurementConfig.setDefaults(self)
104
105 self.plugins.names = ["base_CircularApertureFlux",
106 "base_PixelFlags",
107 "base_SkyCoord",
108 "base_PsfFlux",
109 "base_SdssCentroid",
110 "base_SdssShape",
111 "base_GaussianFlux",
112 "base_PeakLikelihoodFlux",
113 "base_PeakCentroid",
114 "base_NaiveCentroid",
115 "ip_diffim_NaiveDipoleCentroid",
116 "ip_diffim_NaiveDipoleFlux",
117 "ip_diffim_PsfDipoleFlux",
118 "ip_diffim_ClassificationDipole",
119 ]
120
121 self.slots.calibFlux = None
122 self.slots.modelFlux = None
123 self.slots.gaussianFlux = None
124 self.slots.shape = "base_SdssShape"
125 self.slots.centroid = "ip_diffim_NaiveDipoleCentroid"
127
128
129class DipoleFitTask(measBase.SingleFrameMeasurementTask):
130 """A task that fits a dipole to a difference image, with an optional separate detection image.
131
132 Because it subclasses SingleFrameMeasurementTask, and calls
133 SingleFrameMeasurementTask.run() from its run() method, it still
134 can be used identically to a standard SingleFrameMeasurementTask.
135 """
136
137 ConfigClass = DipoleFitTaskConfig
138 _DefaultName = "ip_diffim_DipoleFit"
139
140 def __init__(self, schema, algMetadata=None, **kwargs):
141
142 measBase.SingleFrameMeasurementTask.__init__(self, schema, algMetadata, **kwargs)
143
144 dpFitPluginConfig = self.config.plugins['ip_diffim_DipoleFit']
145
146 self.dipoleFitter = DipoleFitPlugin(dpFitPluginConfig, name=self._DefaultName,
147 schema=schema, metadata=algMetadata,
148 logName=self.log.name)
149
150 @timeMethod
151 def run(self, sources, exposure, posExp=None, negExp=None, **kwargs):
152 """Run dipole measurement and classification
153
154 Parameters
155 ----------
157 ``diaSources`` that will be measured using dipole measurement
158 exposure : `lsst.afw.image.Exposure`
159 The difference exposure on which the ``diaSources`` of the ``sources`` parameter
160 were detected. If neither ``posExp`` nor ``negExp`` are set, then the dipole is also
161 fitted directly to this difference image.
162 posExp : `lsst.afw.image.Exposure`, optional
163 "Positive" exposure, typically a science exposure, or None if unavailable
164 When `posExp` is `None`, will compute `posImage = exposure + negExp`.
165 negExp : `lsst.afw.image.Exposure`, optional
166 "Negative" exposure, typically a template exposure, or None if unavailable
167 When `negExp` is `None`, will compute `negImage = posExp - exposure`.
168 **kwargs
169 Additional keyword arguments for `lsst.meas.base.sfm.SingleFrameMeasurementTask`.
170 """
171
172 measBase.SingleFrameMeasurementTask.run(self, sources, exposure, **kwargs)
173
174 if not sources:
175 return
176
177 for source in sources:
178 self.dipoleFitter.measure(source, exposure, posExp, negExp)
179
180
182 """Lightweight class containing methods for generating a dipole model for fitting
183 to sources in diffims, used by DipoleFitAlgorithm.
184
185 See also:
186 `DMTN-007: Dipole characterization for image differencing <https://dmtn-007.lsst.io>`_.
187 """
188
189 def __init__(self):
190 import lsstDebug
191 self.debug = lsstDebug.Info(__name__).debug
192 self.log = logging.getLogger(__name__)
193
194 def makeBackgroundModel(self, in_x, pars=None):
195 """Generate gradient model (2-d array) with up to 2nd-order polynomial
196
197 Parameters
198 ----------
199 in_x : `numpy.array`
200 (2, w, h)-dimensional `numpy.array`, containing the
201 input x,y meshgrid providing the coordinates upon which to
202 compute the gradient. This will typically be generated via
203 `_generateXYGrid()`. `w` and `h` correspond to the width and
204 height of the desired grid.
205 pars : `list` of `float`, optional
206 Up to 6 floats for up
207 to 6 2nd-order 2-d polynomial gradient parameters, in the
208 following order: (intercept, x, y, xy, x**2, y**2). If `pars`
209 is emtpy or `None`, do nothing and return `None` (for speed).
210
211 Returns
212 -------
213 result : `None` or `numpy.array`
214 return None, or 2-d numpy.array of width/height matching
215 input bbox, containing computed gradient values.
216 """
217
218 # Don't fit for other gradient parameters if the intercept is not included.
219 if (pars is None) or (len(pars) <= 0) or (pars[0] is None):
220 return
221
222 y, x = in_x[0, :], in_x[1, :]
223 gradient = np.full_like(x, pars[0], dtype='float64')
224 if len(pars) > 1 and pars[1] is not None:
225 gradient += pars[1] * x
226 if len(pars) > 2 and pars[2] is not None:
227 gradient += pars[2] * y
228 if len(pars) > 3 and pars[3] is not None:
229 gradient += pars[3] * (x * y)
230 if len(pars) > 4 and pars[4] is not None:
231 gradient += pars[4] * (x * x)
232 if len(pars) > 5 and pars[5] is not None:
233 gradient += pars[5] * (y * y)
234
235 return gradient
236
237 def _generateXYGrid(self, bbox):
238 """Generate a meshgrid covering the x,y coordinates bounded by bbox
239
240 Parameters
241 ----------
242 bbox : `lsst.geom.Box2I`
243 input Bounding Box defining the coordinate limits
244
245 Returns
246 -------
247 in_x : `numpy.array`
248 (2, w, h)-dimensional numpy array containing the grid indexing over x- and
249 y- coordinates
250 """
251
252 x, y = np.mgrid[bbox.getBeginY():bbox.getEndY(), bbox.getBeginX():bbox.getEndX()]
253 in_x = np.array([y, x]).astype(np.float64)
254 in_x[0, :] -= np.mean(in_x[0, :])
255 in_x[1, :] -= np.mean(in_x[1, :])
256 return in_x
257
258 def _getHeavyFootprintSubimage(self, fp, badfill=np.nan, grow=0):
259 """Extract the image from a ``~lsst.afw.detection.HeavyFootprint``
260 as an `lsst.afw.image.ImageF`.
261
262 Parameters
263 ----------
265 HeavyFootprint to use to generate the subimage
266 badfill : `float`, optional
267 Value to fill in pixels in extracted image that are outside the footprint
268 grow : `int`
269 Optionally grow the footprint by this amount before extraction
270
271 Returns
272 -------
273 subim2 : `lsst.afw.image.ImageF`
274 An `~lsst.afw.image.ImageF` containing the subimage.
275 """
276 bbox = fp.getBBox()
277 if grow > 0:
278 bbox.grow(grow)
279
280 subim2 = afwImage.ImageF(bbox, badfill)
281 fp.getSpans().unflatten(subim2.getArray(), fp.getImageArray(), bbox.getCorners()[0])
282 return subim2
283
284 def fitFootprintBackground(self, source, posImage, order=1):
285 """Fit a linear (polynomial) model of given order (max 2) to the background of a footprint.
286
287 Only fit the pixels OUTSIDE of the footprint, but within its bounding box.
288
289 Parameters
290 ----------
292 SourceRecord, the footprint of which is to be fit
293 posImage : `lsst.afw.image.Exposure`
294 The exposure from which to extract the footprint subimage
295 order : `int`
296 Polynomial order of background gradient to fit.
297
298 Returns
299 -------
300 pars : `tuple` of `float`
301 `tuple` of length (1 if order==0; 3 if order==1; 6 if order == 2),
302 containing the resulting fit parameters
303 """
304
305 # TODO look into whether to use afwMath background methods -- see
306 # http://lsst-web.ncsa.illinois.edu/doxygen/x_masterDoxyDoc/_background_example.html
307 fp = source.getFootprint()
308 bbox = fp.getBBox()
309 bbox.grow(3)
310 posImg = afwImage.ImageF(posImage.getMaskedImage().getImage(), bbox, afwImage.PARENT)
311
312 # This code constructs the footprint image so that we can identify the pixels that are
313 # outside the footprint (but within the bounding box). These are the pixels used for
314 # fitting the background.
315 posHfp = afwDet.HeavyFootprintF(fp, posImage.getMaskedImage())
316 posFpImg = self._getHeavyFootprintSubimage(posHfp, grow=3)
317
318 isBg = np.isnan(posFpImg.getArray()).ravel()
319
320 data = posImg.getArray().ravel()
321 data = data[isBg]
322 B = data
323
324 x, y = np.mgrid[bbox.getBeginY():bbox.getEndY(), bbox.getBeginX():bbox.getEndX()]
325 x = x.astype(np.float64).ravel()
326 x -= np.mean(x)
327 x = x[isBg]
328 y = y.astype(np.float64).ravel()
329 y -= np.mean(y)
330 y = y[isBg]
331 b = np.ones_like(x, dtype=np.float64)
332
333 M = np.vstack([b]).T # order = 0
334 if order == 1:
335 M = np.vstack([b, x, y]).T
336 elif order == 2:
337 M = np.vstack([b, x, y, x**2., y**2., x*y]).T
338
339 pars = np.linalg.lstsq(M, B, rcond=-1)[0]
340 return pars
341
342 def makeStarModel(self, bbox, psf, xcen, ycen, flux):
343 """Generate a 2D image model of a single PDF centered at the given coordinates.
344
345 Parameters
346 ----------
347 bbox : `lsst.geom.Box`
348 Bounding box marking pixel coordinates for generated model
349 psf : TODO: DM-17458
350 Psf model used to generate the 'star'
351 xcen : `float`
352 Desired x-centroid of the 'star'
353 ycen : `float`
354 Desired y-centroid of the 'star'
355 flux : `float`
356 Desired flux of the 'star'
357
358 Returns
359 -------
360 p_Im : `lsst.afw.image.Image`
361 2-d stellar image of width/height matching input ``bbox``,
362 containing PSF with given centroid and flux
363 """
364
365 # Generate the psf image, normalize to flux
366 psf_img = psf.computeImage(geom.Point2D(xcen, ycen)).convertF()
367 psf_img_sum = np.nansum(psf_img.getArray())
368 psf_img *= (flux/psf_img_sum)
369
370 # Clip the PSF image bounding box to fall within the footprint bounding box
371 psf_box = psf_img.getBBox()
372 psf_box.clip(bbox)
373 psf_img = afwImage.ImageF(psf_img, psf_box, afwImage.PARENT)
374
375 # Then actually crop the psf image.
376 # Usually not necessary, but if the dipole is near the edge of the image...
377 # Would be nice if we could compare original pos_box with clipped pos_box and
378 # see if it actually was clipped.
379 p_Im = afwImage.ImageF(bbox)
380 tmpSubim = afwImage.ImageF(p_Im, psf_box, afwImage.PARENT)
381 tmpSubim += psf_img
382
383 return p_Im
384
385 def makeModel(self, x, flux, xcenPos, ycenPos, xcenNeg, ycenNeg, fluxNeg=None,
386 b=None, x1=None, y1=None, xy=None, x2=None, y2=None,
387 bNeg=None, x1Neg=None, y1Neg=None, xyNeg=None, x2Neg=None, y2Neg=None,
388 **kwargs):
389 """Generate dipole model with given parameters.
390
391 This is the function whose sum-of-squared difference from data
392 is minimized by `lmfit`.
393
394 x : TODO: DM-17458
395 Input independent variable. Used here as the grid on
396 which to compute the background gradient model.
397 flux : `float`
398 Desired flux of the positive lobe of the dipole
399 xcenPos, ycenPos : `float`
400 Desired x,y-centroid of the positive lobe of the dipole
401 xcenNeg, ycenNeg : `float`
402 Desired x,y-centroid of the negative lobe of the dipole
403 fluxNeg : `float`, optional
404 Desired flux of the negative lobe of the dipole, set to 'flux' if None
405 b, x1, y1, xy, x2, y2 : `float`
406 Gradient parameters for positive lobe.
407 bNeg, x1Neg, y1Neg, xyNeg, x2Neg, y2Neg : `float`, optional
408 Gradient parameters for negative lobe.
409 They are set to the corresponding positive values if None.
410
411 **kwargs : `dict` [`str`]
412 Keyword arguments passed through ``lmfit`` and
413 used by this function. These must include:
414
415 - ``psf`` Psf model used to generate the 'star'
416 - ``rel_weight`` Used to signify least-squares weighting of posImage/negImage
417 relative to diffim. If ``rel_weight == 0`` then posImage/negImage are ignored.
418 - ``bbox`` Bounding box containing region to be modelled
419
420 Returns
421 -------
422 zout : `numpy.array`
423 Has width and height matching the input bbox, and
424 contains the dipole model with given centroids and flux(es). If
425 ``rel_weight`` = 0, this is a 2-d array with dimensions matching
426 those of bbox; otherwise a stack of three such arrays,
427 representing the dipole (diffim), positive, and negative images
428 respectively.
429 """
430
431 psf = kwargs.get('psf')
432 rel_weight = kwargs.get('rel_weight') # if > 0, we're including pre-sub. images
433 fp = kwargs.get('footprint')
434 bbox = fp.getBBox()
435
436 if fluxNeg is None:
437 fluxNeg = flux
438
439 if self.debug:
440 self.log.debug('%.2f %.2f %.2f %.2f %.2f %.2f',
441 flux, fluxNeg, xcenPos, ycenPos, xcenNeg, ycenNeg)
442 if x1 is not None:
443 self.log.debug(' %.2f %.2f %.2f', b, x1, y1)
444 if xy is not None:
445 self.log.debug(' %.2f %.2f %.2f', xy, x2, y2)
446
447 posIm = self.makeStarModel(bbox, psf, xcenPos, ycenPos, flux)
448 negIm = self.makeStarModel(bbox, psf, xcenNeg, ycenNeg, fluxNeg)
449
450 in_x = x
451 if in_x is None: # use the footprint to generate the input grid
452 y, x = np.mgrid[bbox.getBeginY():bbox.getEndY(), bbox.getBeginX():bbox.getEndX()]
453 in_x = np.array([x, y]) * 1.
454 in_x[0, :] -= in_x[0, :].mean() # center it!
455 in_x[1, :] -= in_x[1, :].mean()
456
457 if b is not None:
458 gradient = self.makeBackgroundModel(in_x, (b, x1, y1, xy, x2, y2))
459
460 # If bNeg is None, then don't fit the negative background separately
461 if bNeg is not None:
462 gradientNeg = self.makeBackgroundModel(in_x, (bNeg, x1Neg, y1Neg, xyNeg, x2Neg, y2Neg))
463 else:
464 gradientNeg = gradient
465
466 posIm.getArray()[:, :] += gradient
467 negIm.getArray()[:, :] += gradientNeg
468
469 # Generate the diffIm model
470 diffIm = afwImage.ImageF(bbox)
471 diffIm += posIm
472 diffIm -= negIm
473
474 zout = diffIm.getArray()
475 if rel_weight > 0.:
476 zout = np.append([zout], [posIm.getArray(), negIm.getArray()], axis=0)
477
478 return zout
479
480
482 """Fit a dipole model using an image difference.
483
484 See also:
485 `DMTN-007: Dipole characterization for image differencing <https://dmtn-007.lsst.io>`_.
486 """
487
488 # This is just a private version number to sync with the ipython notebooks that I have been
489 # using for algorithm development.
490 _private_version_ = '0.0.5'
491
492 # Below is a (somewhat incomplete) list of improvements
493 # that would be worth investigating, given the time:
494
495 # todo 1. evaluate necessity for separate parameters for pos- and neg- images
496 # todo 2. only fit background OUTSIDE footprint (DONE) and dipole params INSIDE footprint (NOT DONE)?
497 # todo 3. correct normalization of least-squares weights based on variance planes
498 # todo 4. account for PSFs that vary across the exposures (should be happening by default?)
499 # todo 5. correctly account for NA/masks (i.e., ignore!)
500 # todo 6. better exception handling in the plugin
501 # todo 7. better classification of dipoles (e.g. by comparing chi2 fit vs. monopole?)
502 # todo 8. (DONE) Initial fast estimate of background gradient(s) params -- perhaps using numpy.lstsq
503 # todo 9. (NOT NEEDED - see (2)) Initial fast test whether a background gradient needs to be fit
504 # todo 10. (DONE) better initial estimate for flux when there's a strong gradient
505 # todo 11. (DONE) requires a new package `lmfit` -- investiate others? (astropy/scipy/iminuit?)
506
507 def __init__(self, diffim, posImage=None, negImage=None):
508 """Algorithm to run dipole measurement on a diaSource
509
510 Parameters
511 ----------
512 diffim : `lsst.afw.image.Exposure`
513 Exposure on which the diaSources were detected
514 posImage : `lsst.afw.image.Exposure`
515 "Positive" exposure from which the template was subtracted
516 negImage : `lsst.afw.image.Exposure`
517 "Negative" exposure which was subtracted from the posImage
518 """
519
520 self.diffim = diffim
521 self.posImage = posImage
522 self.negImage = negImage
523 self.psfSigma = None
524 if diffim is not None:
525 diffimPsf = diffim.getPsf()
526 diffimAvgPos = diffimPsf.getAveragePosition()
527 self.psfSigma = diffimPsf.computeShape(diffimAvgPos).getDeterminantRadius()
528
529 self.log = logging.getLogger(__name__)
530
531 import lsstDebug
532 self.debug = lsstDebug.Info(__name__).debug
533
534 def fitDipoleImpl(self, source, tol=1e-7, rel_weight=0.5,
535 fitBackground=1, bgGradientOrder=1, maxSepInSigma=5.,
536 separateNegParams=True, verbose=False):
537 """Fit a dipole model to an input difference image.
538
539 Actually, fits the subimage bounded by the input source's
540 footprint) and optionally constrain the fit using the
541 pre-subtraction images posImage and negImage.
542
543 Parameters
544 ----------
545 source : TODO: DM-17458
546 TODO: DM-17458
547 tol : float, optional
548 TODO: DM-17458
549 rel_weight : `float`, optional
550 TODO: DM-17458
551 fitBackground : `int`, optional
552 TODO: DM-17458
553 bgGradientOrder : `int`, optional
554 TODO: DM-17458
555 maxSepInSigma : `float`, optional
556 TODO: DM-17458
557 separateNegParams : `bool`, optional
558 TODO: DM-17458
559 verbose : `bool`, optional
560 TODO: DM-17458
561
562 Returns
563 -------
564 result : `lmfit.MinimizerResult`
565 return `lmfit.MinimizerResult` object containing the fit
566 parameters and other information.
567 """
568
569 # Only import lmfit if someone wants to use the new DipoleFitAlgorithm.
570 import lmfit
571
572 fp = source.getFootprint()
573 bbox = fp.getBBox()
574 subim = afwImage.MaskedImageF(self.diffim.getMaskedImage(), bbox=bbox, origin=afwImage.PARENT)
575
576 z = diArr = subim.getArrays()[0]
577 weights = 1. / subim.getArrays()[2] # get the weights (=1/variance)
578
579 if rel_weight > 0. and ((self.posImage is not None) or (self.negImage is not None)):
580 if self.negImage is not None:
581 negSubim = afwImage.MaskedImageF(self.negImage.getMaskedImage(), bbox, origin=afwImage.PARENT)
582 if self.posImage is not None:
583 posSubim = afwImage.MaskedImageF(self.posImage.getMaskedImage(), bbox, origin=afwImage.PARENT)
584 if self.posImage is None: # no science image provided; generate it from diffim + negImage
585 posSubim = subim.clone()
586 posSubim += negSubim
587 if self.negImage is None: # no template provided; generate it from the posImage - diffim
588 negSubim = posSubim.clone()
589 negSubim -= subim
590
591 z = np.append([z], [posSubim.getArrays()[0],
592 negSubim.getArrays()[0]], axis=0)
593 # Weight the pos/neg images by rel_weight relative to the diffim
594 weights = np.append([weights], [1. / posSubim.getArrays()[2] * rel_weight,
595 1. / negSubim.getArrays()[2] * rel_weight], axis=0)
596 else:
597 rel_weight = 0. # a short-cut for "don't include the pre-subtraction data"
598
599 # It seems that `lmfit` requires a static functor as its optimized method, which eliminates
600 # the ability to pass a bound method or other class method. Here we write a wrapper which
601 # makes this possible.
602 def dipoleModelFunctor(x, flux, xcenPos, ycenPos, xcenNeg, ycenNeg, fluxNeg=None,
603 b=None, x1=None, y1=None, xy=None, x2=None, y2=None,
604 bNeg=None, x1Neg=None, y1Neg=None, xyNeg=None, x2Neg=None, y2Neg=None,
605 **kwargs):
606 """Generate dipole model with given parameters.
607
608 It simply defers to `modelObj.makeModel()`, where `modelObj` comes
609 out of `kwargs['modelObj']`.
610 """
611 modelObj = kwargs.pop('modelObj')
612 return modelObj.makeModel(x, flux, xcenPos, ycenPos, xcenNeg, ycenNeg, fluxNeg=fluxNeg,
613 b=b, x1=x1, y1=y1, xy=xy, x2=x2, y2=y2,
614 bNeg=bNeg, x1Neg=x1Neg, y1Neg=y1Neg, xyNeg=xyNeg,
615 x2Neg=x2Neg, y2Neg=y2Neg, **kwargs)
616
617 dipoleModel = DipoleModel()
618
619 modelFunctor = dipoleModelFunctor # dipoleModel.makeModel does not work for now.
620 # Create the lmfit model (lmfit uses scipy 'leastsq' option by default - Levenberg-Marquardt)
621 # Note we can also tell it to drop missing values from the data.
622 gmod = lmfit.Model(modelFunctor, verbose=verbose, missing='drop')
623
624 # Add the constraints for centroids, fluxes.
625 # starting constraint - near centroid of footprint
626 fpCentroid = np.array([fp.getCentroid().getX(), fp.getCentroid().getY()])
627 cenNeg = cenPos = fpCentroid
628
629 pks = fp.getPeaks()
630
631 if len(pks) >= 1:
632 cenPos = pks[0].getF() # if individual (merged) peaks were detected, use those
633 if len(pks) >= 2: # peaks are already sorted by centroid flux so take the most negative one
634 cenNeg = pks[-1].getF()
635
636 # For close/faint dipoles the starting locs (min/max) might be way off, let's help them a bit.
637 # First assume dipole is not separated by more than 5*psfSigma.
638 maxSep = self.psfSigma * maxSepInSigma
639
640 # As an initial guess -- assume the dipole is close to the center of the footprint.
641 if np.sum(np.sqrt((np.array(cenPos) - fpCentroid)**2.)) > maxSep:
642 cenPos = fpCentroid
643 if np.sum(np.sqrt((np.array(cenNeg) - fpCentroid)**2.)) > maxSep:
644 cenPos = fpCentroid
645
646 # parameter hints/constraints: https://lmfit.github.io/lmfit-py/model.html#model-param-hints-section
647 # might make sense to not use bounds -- see http://lmfit.github.io/lmfit-py/bounds.html
648 # also see this discussion -- https://github.com/scipy/scipy/issues/3129
649 gmod.set_param_hint('xcenPos', value=cenPos[0],
650 min=cenPos[0]-maxSep, max=cenPos[0]+maxSep)
651 gmod.set_param_hint('ycenPos', value=cenPos[1],
652 min=cenPos[1]-maxSep, max=cenPos[1]+maxSep)
653 gmod.set_param_hint('xcenNeg', value=cenNeg[0],
654 min=cenNeg[0]-maxSep, max=cenNeg[0]+maxSep)
655 gmod.set_param_hint('ycenNeg', value=cenNeg[1],
656 min=cenNeg[1]-maxSep, max=cenNeg[1]+maxSep)
657
658 # Use the (flux under the dipole)*5 for an estimate.
659 # Lots of testing showed that having startingFlux be too high was better than too low.
660 startingFlux = np.nansum(np.abs(diArr) - np.nanmedian(np.abs(diArr))) * 5.
661 posFlux = negFlux = startingFlux
662
663 # TBD: set max. flux limit?
664 gmod.set_param_hint('flux', value=posFlux, min=0.1)
665
666 if separateNegParams:
667 # TBD: set max negative lobe flux limit?
668 gmod.set_param_hint('fluxNeg', value=np.abs(negFlux), min=0.1)
669
670 # Fixed parameters (don't fit for them if there are no pre-sub images or no gradient fit requested):
671 # Right now (fitBackground == 1), we fit a linear model to the background and then subtract
672 # it from the data and then don't fit the background again (this is faster).
673 # A slower alternative (fitBackground == 2) is to use the estimated background parameters as
674 # starting points in the integrated model fit. That is currently not performed by default,
675 # but might be desirable in some cases.
676 bgParsPos = bgParsNeg = (0., 0., 0.)
677 if ((rel_weight > 0.) and (fitBackground != 0) and (bgGradientOrder >= 0)):
678 pbg = 0.
679 bgFitImage = self.posImage if self.posImage is not None else self.negImage
680 # Fit the gradient to the background (linear model)
681 bgParsPos = bgParsNeg = dipoleModel.fitFootprintBackground(source, bgFitImage,
682 order=bgGradientOrder)
683
684 # Generate the gradient and subtract it from the pre-subtraction image data
685 if fitBackground == 1:
686 in_x = dipoleModel._generateXYGrid(bbox)
687 pbg = dipoleModel.makeBackgroundModel(in_x, tuple(bgParsPos))
688 z[1, :] -= pbg
689 z[1, :] -= np.nanmedian(z[1, :])
690 posFlux = np.nansum(z[1, :])
691 gmod.set_param_hint('flux', value=posFlux*1.5, min=0.1)
692
693 if separateNegParams and self.negImage is not None:
694 bgParsNeg = dipoleModel.fitFootprintBackground(source, self.negImage,
695 order=bgGradientOrder)
696 pbg = dipoleModel.makeBackgroundModel(in_x, tuple(bgParsNeg))
697 z[2, :] -= pbg
698 z[2, :] -= np.nanmedian(z[2, :])
699 if separateNegParams:
700 negFlux = np.nansum(z[2, :])
701 gmod.set_param_hint('fluxNeg', value=negFlux*1.5, min=0.1)
702
703 # Do not subtract the background from the images but include the background parameters in the fit
704 if fitBackground == 2:
705 if bgGradientOrder >= 0:
706 gmod.set_param_hint('b', value=bgParsPos[0])
707 if separateNegParams:
708 gmod.set_param_hint('bNeg', value=bgParsNeg[0])
709 if bgGradientOrder >= 1:
710 gmod.set_param_hint('x1', value=bgParsPos[1])
711 gmod.set_param_hint('y1', value=bgParsPos[2])
712 if separateNegParams:
713 gmod.set_param_hint('x1Neg', value=bgParsNeg[1])
714 gmod.set_param_hint('y1Neg', value=bgParsNeg[2])
715 if bgGradientOrder >= 2:
716 gmod.set_param_hint('xy', value=bgParsPos[3])
717 gmod.set_param_hint('x2', value=bgParsPos[4])
718 gmod.set_param_hint('y2', value=bgParsPos[5])
719 if separateNegParams:
720 gmod.set_param_hint('xyNeg', value=bgParsNeg[3])
721 gmod.set_param_hint('x2Neg', value=bgParsNeg[4])
722 gmod.set_param_hint('y2Neg', value=bgParsNeg[5])
723
724 y, x = np.mgrid[bbox.getBeginY():bbox.getEndY(), bbox.getBeginX():bbox.getEndX()]
725 in_x = np.array([x, y]).astype(np.float64)
726 in_x[0, :] -= in_x[0, :].mean() # center it!
727 in_x[1, :] -= in_x[1, :].mean()
728
729 # Instead of explicitly using a mask to ignore flagged pixels, just set the ignored pixels'
730 # weights to 0 in the fit. TBD: need to inspect mask planes to set this mask.
731 mask = np.ones_like(z, dtype=bool) # TBD: set mask values to False if the pixels are to be ignored
732
733 # I'm not sure about the variance planes in the diffim (or convolved pre-sub. images
734 # for that matter) so for now, let's just do an un-weighted least-squares fit
735 # (override weights computed above).
736 weights = mask.astype(np.float64)
737 if self.posImage is not None and rel_weight > 0.:
738 weights = np.array([np.ones_like(diArr), np.ones_like(diArr)*rel_weight,
739 np.ones_like(diArr)*rel_weight])
740
741 # Set the weights to zero if mask is False
742 if np.any(~mask):
743 weights[~mask] = 0.
744
745 # Note that although we can, we're not required to set initial values for params here,
746 # since we set their param_hint's above.
747 # Can add "method" param to not use 'leastsq' (==levenberg-marquardt), e.g. "method='nelder'"
748 with warnings.catch_warnings():
749 # Ignore lmfit unknown argument warnings:
750 # "psf, rel_weight, footprint, modelObj" all become pass-through kwargs for makeModel.
751 warnings.filterwarnings("ignore", "The keyword argument .* does not match", UserWarning)
752 result = gmod.fit(z, weights=weights, x=in_x, max_nfev=250,
753 method="leastsq", # TODO: try using `least_squares` here for speed/robustness
754 verbose=verbose,
755 # see scipy docs for the meaning of these keywords
756 fit_kws={'ftol': tol, 'xtol': tol, 'gtol': tol,
757 # Our model is float32 internally, so we need a larger epsfcn.
758 'epsfcn': 1e-10},
759 psf=self.diffim.getPsf(), # hereon: kwargs that get passed to makeModel()
760 rel_weight=rel_weight,
761 footprint=fp,
762 modelObj=dipoleModel)
763
764 if verbose: # the ci_report() seems to fail if neg params are constrained -- TBD why.
765 # Never wanted in production - this takes a long time (longer than the fit!)
766 # This is how to get confidence intervals out:
767 # https://lmfit.github.io/lmfit-py/confidence.html and
768 # http://cars9.uchicago.edu/software/python/lmfit/model.html
769 print(result.fit_report(show_correl=False))
770 if separateNegParams:
771 print(result.ci_report())
772
773 return result
774
775 def fitDipole(self, source, tol=1e-7, rel_weight=0.1,
776 fitBackground=1, maxSepInSigma=5., separateNegParams=True,
777 bgGradientOrder=1, verbose=False, display=False):
778 """Fit a dipole model to an input ``diaSource`` (wraps `fitDipoleImpl`).
779
780 Actually, fits the subimage bounded by the input source's
781 footprint) and optionally constrain the fit using the
782 pre-subtraction images self.posImage (science) and
783 self.negImage (template). Wraps the output into a
784 `pipeBase.Struct` named tuple after computing additional
785 statistics such as orientation and SNR.
786
787 Parameters
788 ----------
790 Record containing the (merged) dipole source footprint detected on the diffim
791 tol : `float`, optional
792 Tolerance parameter for scipy.leastsq() optimization
793 rel_weight : `float`, optional
794 Weighting of posImage/negImage relative to the diffim in the fit
795 fitBackground : `int`, {0, 1, 2}, optional
796 How to fit linear background gradient in posImage/negImage
797
798 - 0: do not fit background at all
799 - 1 (default): pre-fit the background using linear least squares and then do not fit it
800 as part of the dipole fitting optimization
801 - 2: pre-fit the background using linear least squares (as in 1), and use the parameter
802 estimates from that fit as starting parameters for an integrated "re-fit" of the
803 background as part of the overall dipole fitting optimization.
804 maxSepInSigma : `float`, optional
805 Allowed window of centroid parameters relative to peak in input source footprint
806 separateNegParams : `bool`, optional
807 Fit separate parameters to the flux and background gradient in
808 bgGradientOrder : `int`, {0, 1, 2}, optional
809 Desired polynomial order of background gradient
810 verbose: `bool`, optional
811 Be verbose
812 display
813 Display input data, best fit model(s) and residuals in a matplotlib window.
814
815 Returns
816 -------
817 result : `struct`
818 `pipeBase.Struct` object containing the fit parameters and other information.
819
820 result : `callable`
821 `lmfit.MinimizerResult` object for debugging and error estimation, etc.
822
823 Notes
824 -----
825 Parameter `fitBackground` has three options, thus it is an integer:
826
827 """
828
829 fitResult = self.fitDipoleImpl(
830 source, tol=tol, rel_weight=rel_weight, fitBackground=fitBackground,
831 maxSepInSigma=maxSepInSigma, separateNegParams=separateNegParams,
832 bgGradientOrder=bgGradientOrder, verbose=verbose)
833
834 # Display images, model fits and residuals (currently uses matplotlib display functions)
835 if display:
836 fp = source.getFootprint()
837 self.displayFitResults(fp, fitResult)
838
839 fitParams = fitResult.best_values
840 if fitParams['flux'] <= 1.: # usually around 0.1 -- the minimum flux allowed -- i.e. bad fit.
841 out = Struct(posCentroidX=np.nan, posCentroidY=np.nan,
842 negCentroidX=np.nan, negCentroidY=np.nan,
843 posFlux=np.nan, negFlux=np.nan, posFluxErr=np.nan, negFluxErr=np.nan,
844 centroidX=np.nan, centroidY=np.nan, orientation=np.nan,
845 signalToNoise=np.nan, chi2=np.nan, redChi2=np.nan)
846 return out, fitResult
847
848 centroid = ((fitParams['xcenPos'] + fitParams['xcenNeg']) / 2.,
849 (fitParams['ycenPos'] + fitParams['ycenNeg']) / 2.)
850 dx, dy = fitParams['xcenPos'] - fitParams['xcenNeg'], fitParams['ycenPos'] - fitParams['ycenNeg']
851 angle = np.arctan2(dy, dx) / np.pi * 180. # convert to degrees (should keep as rad?)
852
853 # Exctract flux value, compute signalToNoise from flux/variance_within_footprint
854 # Also extract the stderr of flux estimate.
855 def computeSumVariance(exposure, footprint):
856 box = footprint.getBBox()
857 subim = afwImage.MaskedImageF(exposure.getMaskedImage(), box, origin=afwImage.PARENT)
858 return np.sqrt(np.nansum(subim.getArrays()[1][:, :]))
859
860 fluxVal = fluxVar = fitParams['flux']
861 fluxErr = fluxErrNeg = fitResult.params['flux'].stderr
862 if self.posImage is not None:
863 fluxVar = computeSumVariance(self.posImage, source.getFootprint())
864 else:
865 fluxVar = computeSumVariance(self.diffim, source.getFootprint())
866
867 fluxValNeg, fluxVarNeg = fluxVal, fluxVar
868 if separateNegParams:
869 fluxValNeg = fitParams['fluxNeg']
870 fluxErrNeg = fitResult.params['fluxNeg'].stderr
871 if self.negImage is not None:
872 fluxVarNeg = computeSumVariance(self.negImage, source.getFootprint())
873
874 try:
875 signalToNoise = np.sqrt((fluxVal/fluxVar)**2 + (fluxValNeg/fluxVarNeg)**2)
876 except ZeroDivisionError: # catch divide by zero - should never happen.
877 signalToNoise = np.nan
878
879 out = Struct(posCentroidX=fitParams['xcenPos'], posCentroidY=fitParams['ycenPos'],
880 negCentroidX=fitParams['xcenNeg'], negCentroidY=fitParams['ycenNeg'],
881 posFlux=fluxVal, negFlux=-fluxValNeg, posFluxErr=fluxErr, negFluxErr=fluxErrNeg,
882 centroidX=centroid[0], centroidY=centroid[1], orientation=angle,
883 signalToNoise=signalToNoise, chi2=fitResult.chisqr, redChi2=fitResult.redchi)
884
885 # fitResult may be returned for debugging
886 return out, fitResult
887
888 def displayFitResults(self, footprint, result):
889 """Display data, model fits and residuals (currently uses matplotlib display functions).
890
891 Parameters
892 ----------
893 footprint : TODO: DM-17458
894 Footprint containing the dipole that was fit
895 result : `lmfit.MinimizerResult`
896 `lmfit.MinimizerResult` object returned by `lmfit` optimizer
897
898 Returns
899 -------
900 fig : `matplotlib.pyplot.plot`
901 """
902 try:
903 import matplotlib.pyplot as plt
904 except ImportError as err:
905 self.log.warning('Unable to import matplotlib: %s', err)
906 raise err
907
908 def display2dArray(arr, title='Data', extent=None):
909 """Use `matplotlib.pyplot.imshow` to display a 2-D array with a given coordinate range.
910 """
911 fig = plt.imshow(arr, origin='lower', interpolation='none', cmap='gray', extent=extent)
912 plt.title(title)
913 plt.colorbar(fig, cmap='gray')
914 return fig
915
916 z = result.data
917 fit = result.best_fit
918 bbox = footprint.getBBox()
919 extent = (bbox.getBeginX(), bbox.getEndX(), bbox.getBeginY(), bbox.getEndY())
920 if z.shape[0] == 3:
921 fig = plt.figure(figsize=(8, 8))
922 for i in range(3):
923 plt.subplot(3, 3, i*3+1)
924 display2dArray(z[i, :], 'Data', extent=extent)
925 plt.subplot(3, 3, i*3+2)
926 display2dArray(fit[i, :], 'Model', extent=extent)
927 plt.subplot(3, 3, i*3+3)
928 display2dArray(z[i, :] - fit[i, :], 'Residual', extent=extent)
929 return fig
930 else:
931 fig = plt.figure(figsize=(8, 2.5))
932 plt.subplot(1, 3, 1)
933 display2dArray(z, 'Data', extent=extent)
934 plt.subplot(1, 3, 2)
935 display2dArray(fit, 'Model', extent=extent)
936 plt.subplot(1, 3, 3)
937 display2dArray(z - fit, 'Residual', extent=extent)
938 return fig
939
940 plt.show()
941
942
943@measBase.register("ip_diffim_DipoleFit")
944class DipoleFitPlugin(measBase.SingleFramePlugin):
945 """A single frame measurement plugin that fits dipoles to all merged (two-peak) ``diaSources``.
946
947 This measurement plugin accepts up to three input images in
948 its `measure` method. If these are provided, it includes data
949 from the pre-subtraction posImage (science image) and optionally
950 negImage (template image) to constrain the fit. The meat of the
951 fitting routines are in the class `~lsst.module.name.DipoleFitAlgorithm`.
952
953 Notes
954 -----
955 The motivation behind this plugin and the necessity for including more than
956 one exposure are documented in DMTN-007 (http://dmtn-007.lsst.io).
957
958 This class is named `ip_diffim_DipoleFit` so that it may be used alongside
959 the existing `ip_diffim_DipoleMeasurement` classes until such a time as those
960 are deemed to be replaceable by this.
961 """
962
963 ConfigClass = DipoleFitPluginConfig
964 DipoleFitAlgorithmClass = DipoleFitAlgorithm # Pointer to the class that performs the fit
965
966 FAILURE_EDGE = 1 # too close to the edge
967 FAILURE_FIT = 2 # failure in the fitting
968 FAILURE_NOT_DIPOLE = 4 # input source is not a putative dipole to begin with
969
970 @classmethod
972 """Set execution order to `FLUX_ORDER`.
973
974 This includes algorithms that require both `getShape()` and `getCentroid()`,
975 in addition to a Footprint and its Peaks.
976 """
977 return cls.FLUX_ORDER
978
979 def __init__(self, config, name, schema, metadata, logName=None):
980 if logName is None:
981 logName = name
982 measBase.SingleFramePlugin.__init__(self, config, name, schema, metadata, logName=logName)
983
984 self.log = logging.getLogger(logName)
985
986 self._setupSchema(config, name, schema, metadata)
987
988 def _setupSchema(self, config, name, schema, metadata):
989 # Get a FunctorKey that can quickly look up the "blessed" centroid value.
990 self.centroidKey = afwTable.Point2DKey(schema["slot_Centroid"])
991
992 # Add some fields for our outputs, and save their Keys.
993 # Use setattr() to programmatically set the pos/neg named attributes to values, e.g.
994 # self.posCentroidKeyX = 'ip_diffim_DipoleFit_pos_centroid_x'
995
996 for pos_neg in ['pos', 'neg']:
997
998 key = schema.addField(
999 schema.join(name, pos_neg, "instFlux"), type=float, units="count",
1000 doc="Dipole {0} lobe flux".format(pos_neg))
1001 setattr(self, ''.join((pos_neg, 'FluxKey')), key)
1002
1003 key = schema.addField(
1004 schema.join(name, pos_neg, "instFluxErr"), type=float, units="count",
1005 doc="1-sigma uncertainty for {0} dipole flux".format(pos_neg))
1006 setattr(self, ''.join((pos_neg, 'FluxErrKey')), key)
1007
1008 for x_y in ['x', 'y']:
1009 key = schema.addField(
1010 schema.join(name, pos_neg, "centroid", x_y), type=float, units="pixel",
1011 doc="Dipole {0} lobe centroid".format(pos_neg))
1012 setattr(self, ''.join((pos_neg, 'CentroidKey', x_y.upper())), key)
1013
1014 for x_y in ['x', 'y']:
1015 key = schema.addField(
1016 schema.join(name, "centroid", x_y), type=float, units="pixel",
1017 doc="Dipole centroid")
1018 setattr(self, ''.join(('centroidKey', x_y.upper())), key)
1019
1020 self.fluxKey = schema.addField(
1021 schema.join(name, "instFlux"), type=float, units="count",
1022 doc="Dipole overall flux")
1023
1024 self.orientationKey = schema.addField(
1025 schema.join(name, "orientation"), type=float, units="deg",
1026 doc="Dipole orientation")
1027
1028 self.separationKey = schema.addField(
1029 schema.join(name, "separation"), type=float, units="pixel",
1030 doc="Pixel separation between positive and negative lobes of dipole")
1031
1032 self.chi2dofKey = schema.addField(
1033 schema.join(name, "chi2dof"), type=float,
1034 doc="Chi2 per degree of freedom of dipole fit")
1035
1036 self.signalToNoiseKey = schema.addField(
1037 schema.join(name, "signalToNoise"), type=float,
1038 doc="Estimated signal-to-noise of dipole fit")
1039
1040 self.classificationFlagKey = schema.addField(
1041 schema.join(name, "flag", "classification"), type="Flag",
1042 doc="Flag indicating diaSource is classified as a dipole")
1043
1044 self.classificationAttemptedFlagKey = schema.addField(
1045 schema.join(name, "flag", "classificationAttempted"), type="Flag",
1046 doc="Flag indicating diaSource was attempted to be classified as a dipole")
1047
1048 self.flagKey = schema.addField(
1049 schema.join(name, "flag"), type="Flag",
1050 doc="General failure flag for dipole fit")
1051
1052 self.edgeFlagKey = schema.addField(
1053 schema.join(name, "flag", "edge"), type="Flag",
1054 doc="Flag set when dipole is too close to edge of image")
1055
1056 def measure(self, measRecord, exposure, posExp=None, negExp=None):
1057 """Perform the non-linear least squares minimization on the putative dipole source.
1058
1059 Parameters
1060 ----------
1061 measRecord : `lsst.afw.table.SourceRecord`
1062 diaSources that will be measured using dipole measurement
1063 exposure : `lsst.afw.image.Exposure`
1064 Difference exposure on which the diaSources were detected; `exposure = posExp-negExp`
1065 If both `posExp` and `negExp` are `None`, will attempt to fit the
1066 dipole to just the `exposure` with no constraint.
1067 posExp : `lsst.afw.image.Exposure`, optional
1068 "Positive" exposure, typically a science exposure, or None if unavailable
1069 When `posExp` is `None`, will compute `posImage = exposure + negExp`.
1070 negExp : `lsst.afw.image.Exposure`, optional
1071 "Negative" exposure, typically a template exposure, or None if unavailable
1072 When `negExp` is `None`, will compute `negImage = posExp - exposure`.
1073
1074 Notes
1075 -----
1076 The main functionality of this routine was placed outside of
1077 this plugin (into `DipoleFitAlgorithm.fitDipole()`) so that
1078 `DipoleFitAlgorithm.fitDipole()` can be called separately for
1079 testing (@see `tests/testDipoleFitter.py`)
1080
1081 Returns
1082 -------
1083 result : TODO: DM-17458
1084 TODO: DM-17458
1085 """
1086
1087 result = None
1088 pks = measRecord.getFootprint().getPeaks()
1089
1090 # Check if the footprint consists of a putative dipole - else don't fit it.
1091 if (
1092 (len(pks) <= 1) # one peak in the footprint - not a dipole
1093 or (len(pks) > 1 and (np.sign(pks[0].getPeakValue())
1094 == np.sign(pks[-1].getPeakValue()))) # peaks are same sign - not a dipole
1095 ):
1096 measRecord.set(self.classificationFlagKey, False)
1097 measRecord.set(self.classificationAttemptedFlagKey, False)
1098 self.fail(measRecord, measBase.MeasurementError('not a dipole', self.FAILURE_NOT_DIPOLE))
1099 if not self.config.fitAllDiaSources:
1100 return result
1101
1102 try:
1103 alg = self.DipoleFitAlgorithmClass(exposure, posImage=posExp, negImage=negExp)
1104 result, _ = alg.fitDipole(
1105 measRecord, rel_weight=self.config.relWeight,
1106 tol=self.config.tolerance,
1107 maxSepInSigma=self.config.maxSeparation,
1108 fitBackground=self.config.fitBackground,
1109 separateNegParams=self.config.fitSeparateNegParams,
1110 verbose=False, display=False)
1111 except pexExcept.LengthError:
1112 self.fail(measRecord, measBase.MeasurementError('edge failure', self.FAILURE_EDGE))
1113 except Exception as e:
1114 self.fail(measRecord, measBase.MeasurementError('Exception in dipole fit', self.FAILURE_FIT))
1115 self.log.error("Exception in dipole fit. %s: %s", e.__class__.__name__, e)
1116
1117 if result is None:
1118 measRecord.set(self.classificationFlagKey, False)
1119 measRecord.set(self.classificationAttemptedFlagKey, False)
1120 return result
1121
1122 self.log.debug("Dipole fit result: %d %s", measRecord.getId(), str(result))
1123
1124 if result.posFlux <= 1.: # usually around 0.1 -- the minimum flux allowed -- i.e. bad fit.
1125 self.fail(measRecord, measBase.MeasurementError('dipole fit failure', self.FAILURE_FIT))
1126
1127 # add chi2, coord/flux uncertainties (TBD), dipole classification
1128 # Add the relevant values to the measRecord
1129 measRecord[self.posFluxKey] = result.posFlux
1130 measRecord[self.posFluxErrKey] = result.signalToNoise # to be changed to actual sigma!
1131 measRecord[self.posCentroidKeyX] = result.posCentroidX
1132 measRecord[self.posCentroidKeyY] = result.posCentroidY
1133
1134 measRecord[self.negFluxKey] = result.negFlux
1135 measRecord[self.negFluxErrKey] = result.signalToNoise # to be changed to actual sigma!
1136 measRecord[self.negCentroidKeyX] = result.negCentroidX
1137 measRecord[self.negCentroidKeyY] = result.negCentroidY
1138
1139 # Dia source flux: average of pos+neg
1140 measRecord[self.fluxKey] = (abs(result.posFlux) + abs(result.negFlux))/2.
1141 measRecord[self.orientationKey] = result.orientation
1142 measRecord[self.separationKey] = np.sqrt((result.posCentroidX - result.negCentroidX)**2.
1143 + (result.posCentroidY - result.negCentroidY)**2.)
1144 measRecord[self.centroidKeyX] = result.centroidX
1145 measRecord[self.centroidKeyY] = result.centroidY
1146
1147 measRecord[self.signalToNoiseKey] = result.signalToNoise
1148 measRecord[self.chi2dofKey] = result.redChi2
1149
1150 self.doClassify(measRecord, result.chi2)
1151
1152 def doClassify(self, measRecord, chi2val):
1153 """Classify a source as a dipole.
1154
1155 Parameters
1156 ----------
1157 measRecord : TODO: DM-17458
1158 TODO: DM-17458
1159 chi2val : TODO: DM-17458
1160 TODO: DM-17458
1161
1162 Notes
1163 -----
1164 Sources are classified as dipoles, or not, according to three criteria:
1165
1166 1. Does the total signal-to-noise surpass the ``minSn``?
1167 2. Are the pos/neg fluxes greater than 1.0 and no more than 0.65 (``maxFluxRatio``)
1168 of the total flux? By default this will never happen since ``posFlux == negFlux``.
1169 3. Is it a good fit (``chi2dof`` < 1)? (Currently not used.)
1170 """
1171
1172 # First, does the total signal-to-noise surpass the minSn?
1173 passesSn = measRecord[self.signalToNoiseKey] > self.config.minSn
1174
1175 # Second, are the pos/neg fluxes greater than 1.0 and no more than 0.65 (param maxFluxRatio)
1176 # of the total flux? By default this will never happen since posFlux = negFlux.
1177 passesFluxPos = (abs(measRecord[self.posFluxKey])
1178 / (measRecord[self.fluxKey]*2.)) < self.config.maxFluxRatio
1179 passesFluxPos &= (abs(measRecord[self.posFluxKey]) >= 1.0)
1180 passesFluxNeg = (abs(measRecord[self.negFluxKey])
1181 / (measRecord[self.fluxKey]*2.)) < self.config.maxFluxRatio
1182 passesFluxNeg &= (abs(measRecord[self.negFluxKey]) >= 1.0)
1183 allPass = (passesSn and passesFluxPos and passesFluxNeg) # and passesChi2)
1184
1185 # Third, is it a good fit (chi2dof < 1)?
1186 # Use scipy's chi2 cumulative distrib to estimate significance
1187 # This doesn't really work since I don't trust the values in the variance plane (which
1188 # affects the least-sq weights, which affects the resulting chi2).
1189 # But I'm going to keep this here for future use.
1190 if False:
1191 from scipy.stats import chi2
1192 ndof = chi2val / measRecord[self.chi2dofKey]
1193 significance = chi2.cdf(chi2val, ndof)
1194 passesChi2 = significance < self.config.maxChi2DoF
1195 allPass = allPass and passesChi2
1196
1197 measRecord.set(self.classificationAttemptedFlagKey, True)
1198
1199 if allPass: # Note cannot pass `allPass` into the `measRecord.set()` call below...?
1200 measRecord.set(self.classificationFlagKey, True)
1201 else:
1202 measRecord.set(self.classificationFlagKey, False)
1203
1204 def fail(self, measRecord, error=None):
1205 """Catch failures and set the correct flags.
1206 """
1207
1208 measRecord.set(self.flagKey, True)
1209 if error is not None:
1210 if error.getFlagBit() == self.FAILURE_EDGE:
1211 self.log.warning('DipoleFitPlugin not run on record %d: %s', measRecord.getId(), str(error))
1212 measRecord.set(self.edgeFlagKey, True)
1213 if error.getFlagBit() == self.FAILURE_FIT:
1214 self.log.warning('DipoleFitPlugin failed on record %d: %s', measRecord.getId(), str(error))
1215 measRecord.set(self.flagKey, True)
1216 if error.getFlagBit() == self.FAILURE_NOT_DIPOLE:
1217 self.log.debug('DipoleFitPlugin not run on record %d: %s',
1218 measRecord.getId(), str(error))
1219 measRecord.set(self.classificationAttemptedFlagKey, False)
1220 measRecord.set(self.flagKey, True)
1221 else:
1222 self.log.warning('DipoleFitPlugin failed on record %d', measRecord.getId())
def fitDipoleImpl(self, source, tol=1e-7, rel_weight=0.5, fitBackground=1, bgGradientOrder=1, maxSepInSigma=5., separateNegParams=True, verbose=False)
def fitDipole(self, source, tol=1e-7, rel_weight=0.1, fitBackground=1, maxSepInSigma=5., separateNegParams=True, bgGradientOrder=1, verbose=False, display=False)
def displayFitResults(self, footprint, result)
def __init__(self, diffim, posImage=None, negImage=None)
def fail(self, measRecord, error=None)
def measure(self, measRecord, exposure, posExp=None, negExp=None)
def doClassify(self, measRecord, chi2val)
def _setupSchema(self, config, name, schema, metadata)
def __init__(self, config, name, schema, metadata, logName=None)
def __init__(self, schema, algMetadata=None, **kwargs)
def run(self, sources, exposure, posExp=None, negExp=None, **kwargs)
def _getHeavyFootprintSubimage(self, fp, badfill=np.nan, grow=0)
def makeModel(self, x, flux, xcenPos, ycenPos, xcenNeg, ycenNeg, fluxNeg=None, b=None, x1=None, y1=None, xy=None, x2=None, y2=None, bNeg=None, x1Neg=None, y1Neg=None, xyNeg=None, x2Neg=None, y2Neg=None, **kwargs)
def makeBackgroundModel(self, in_x, pars=None)
def fitFootprintBackground(self, source, posImage, order=1)
def makeStarModel(self, bbox, psf, xcen, ycen, flux)