LSST Applications  21.0.0+04719a4bac,21.0.0-1-ga51b5d4+f5e6047307,21.0.0-11-g2b59f77+a9c1acf22d,21.0.0-11-ga42c5b2+86977b0b17,21.0.0-12-gf4ce030+76814010d2,21.0.0-13-g1721dae+760e7a6536,21.0.0-13-g3a573fe+768d78a30a,21.0.0-15-g5a7caf0+f21cbc5713,21.0.0-16-g0fb55c1+b60e2d390c,21.0.0-19-g4cded4ca+71a93a33c0,21.0.0-2-g103fe59+bb20972958,21.0.0-2-g45278ab+04719a4bac,21.0.0-2-g5242d73+3ad5d60fb1,21.0.0-2-g7f82c8f+8babb168e8,21.0.0-2-g8f08a60+06509c8b61,21.0.0-2-g8faa9b5+616205b9df,21.0.0-2-ga326454+8babb168e8,21.0.0-2-gde069b7+5e4aea9c2f,21.0.0-2-gecfae73+1d3a86e577,21.0.0-2-gfc62afb+3ad5d60fb1,21.0.0-25-g1d57be3cd+e73869a214,21.0.0-3-g357aad2+ed88757d29,21.0.0-3-g4a4ce7f+3ad5d60fb1,21.0.0-3-g4be5c26+3ad5d60fb1,21.0.0-3-g65f322c+e0b24896a3,21.0.0-3-g7d9da8d+616205b9df,21.0.0-3-ge02ed75+a9c1acf22d,21.0.0-4-g591bb35+a9c1acf22d,21.0.0-4-g65b4814+b60e2d390c,21.0.0-4-gccdca77+0de219a2bc,21.0.0-4-ge8a399c+6c55c39e83,21.0.0-5-gd00fb1e+05fce91b99,21.0.0-6-gc675373+3ad5d60fb1,21.0.0-64-g1122c245+4fb2b8f86e,21.0.0-7-g04766d7+cd19d05db2,21.0.0-7-gdf92d54+04719a4bac,21.0.0-8-g5674e7b+d1bd76f71f,master-gac4afde19b+a9c1acf22d,w.2021.13
LSST Data Management Base Package
plotPtc.py
Go to the documentation of this file.
1 # This file is part of cp_pipe.
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 #
22 
23 __all__ = ['PlotPhotonTransferCurveTask']
24 
25 import numpy as np
26 import matplotlib.pyplot as plt
27 import matplotlib as mpl
28 from matplotlib import gridspec
29 import os
30 from matplotlib.backends.backend_pdf import PdfPages
31 
32 import lsst.ip.isr as isr
33 
34 from lsst.cp.pipe.utils import (funcAstier, funcPolynomial,
35  calculateWeightedReducedChi2)
36 from matplotlib.ticker import MaxNLocator
37 from lsst.cp.pipe.ptc.astierCovPtcUtils import getFitDataFromCovariances
38 from lsst.ip.isr import PhotonTransferCurveDataset
39 import lsst.log as lsstLog
40 
41 
43  """A class to plot the dataset from MeasurePhotonTransferCurveTask.
44 
45  Parameters
46  ----------
47 
48  datasetFileName : `str`
49  datasetPtc (lsst.ip.isr.PhotonTransferCurveDataset) file
50  name (fits).
51 
52  linearizerFileName : `str`, optional
53  linearizer (isr.linearize.Linearizer) file
54  name (fits).
55 
56  outDir : `str`, optional
57  Path to the output directory where the final PDF will
58  be placed.
59 
60  detNum : `int`, optional
61  Detector number.
62 
63  signalElectronsRelativeA : `float`, optional
64  Signal value for relative systematic bias between different
65  methods of estimating a_ij (Fig. 15 of Astier+19).
66 
67  plotNormalizedCovariancesNumberOfBins : `float`, optional
68  Number of bins in `plotNormalizedCovariancesNumber` function
69  (Fig. 8, 10., of Astier+19).
70 
71  Notes
72  -----
73  The plotting code in this file is almost identical to the code in
74  `plotPtcGen2.py`. If further changes are implemented in this file,
75  `plotPtcGen2.py` needs to be updated accordingly, and vice versa.
76  The file `plotPtcGen2.py` helps with maintaining backwards
77  compatibility with gen2 as we transition to gen3; the code
78  duplication is meant to only last for few month from now
79  (Jan, 2021). At that point only this file, `plotPtc.py`, will
80  remain.
81  """
82 
83  def __init__(self, datasetFilename, linearizerFileName=None,
84  outDir='.', detNum=999, signalElectronsRelativeA=75000,
85  plotNormalizedCovariancesNumberOfBins=10):
86  self.datasetFilenamedatasetFilename = datasetFilename
87  self.linearizerFileNamelinearizerFileName = linearizerFileName
88  self.detNumdetNum = detNum
89  self.signalElectronsRelativeAsignalElectronsRelativeA = signalElectronsRelativeA
90  self.plotNormalizedCovariancesNumberOfBinsplotNormalizedCovariancesNumberOfBins = plotNormalizedCovariancesNumberOfBins
91  self.outDiroutDir = outDir
92 
93  def runDataRef(self):
94  """Run the Photon Transfer Curve (PTC) plotting measurement task.
95  """
96 
97  datasetFile = self.datasetFilenamedatasetFilename
98  datasetPtc = PhotonTransferCurveDataset.readFits(datasetFile)
99 
100  dirname = self.outDiroutDir
101  if not os.path.exists(dirname):
102  os.makedirs(dirname)
103 
104  detNum = self.detNumdetNum
105  filename = f"PTC_det{detNum}.pdf"
106  filenameFull = os.path.join(dirname, filename)
107 
108  if self.linearizerFileNamelinearizerFileName:
109  linearizer = isr.linearize.Linearizer.readFits(self.linearizerFileNamelinearizerFileName)
110  else:
111  linearizer = None
112  self.runrun(filenameFull, datasetPtc, linearizer=linearizer, log=lsstLog)
113 
114  return
115 
116  def run(self, filenameFull, datasetPtc, linearizer=None, log=None):
117  """Make the plots for the PTC task"""
118  ptcFitType = datasetPtc.ptcFitType
119  with PdfPages(filenameFull) as pdfPages:
120  if ptcFitType in ["FULLCOVARIANCE", ]:
121  self.covAstierMakeAllPlotscovAstierMakeAllPlots(datasetPtc, pdfPages, log=log)
122  elif ptcFitType in ["EXPAPPROXIMATION", "POLYNOMIAL"]:
123  self._plotStandardPtc_plotStandardPtc(datasetPtc, ptcFitType, pdfPages)
124  else:
125  raise RuntimeError(f"The input dataset had an invalid dataset.ptcFitType: {ptcFitType}. \n"
126  "Options: 'FULLCOVARIANCE', EXPAPPROXIMATION, or 'POLYNOMIAL'.")
127  if linearizer:
128  self._plotLinearizer_plotLinearizer(datasetPtc, linearizer, pdfPages)
129 
130  return
131 
132  def covAstierMakeAllPlots(self, dataset, pdfPages,
133  log=None):
134  """Make plots for MeasurePhotonTransferCurve task when doCovariancesAstier=True.
135 
136  This function call other functions that mostly reproduce the plots in Astier+19.
137  Most of the code is ported from Pierre Astier's repository https://github.com/PierreAstier/bfptc
138 
139  Parameters
140  ----------
141  dataset : `lsst.ip.isr.ptcDataset.PhotonTransferCurveDataset`
142  The dataset containing the necessary information to produce the plots.
143 
144  pdfPages: `matplotlib.backends.backend_pdf.PdfPages`
145  PDF file where the plots will be saved.
146 
147  log : `lsst.log.Log`, optional
148  Logger to handle messages
149  """
150  mu = dataset.finalMeans
151  # dictionaries with ampNames as keys
152  fullCovs = dataset.covariances
153  fullCovsModel = dataset.covariancesModel
154  fullCovWeights = dataset.covariancesSqrtWeights
155  aDict = dataset.aMatrix
156  bDict = dataset.bMatrix
157  fullCovsNoB = dataset.covariances
158  fullCovsModelNoB = dataset.covariancesModelNoB
159  fullCovWeightsNoB = dataset.covariancesSqrtWeights
160  aDictNoB = dataset.aMatrixNoB
161  gainDict = dataset.gain
162  noiseDict = dataset.noise
163 
164  self.plotCovariancesplotCovariances(mu, fullCovs, fullCovsModel, fullCovWeights, fullCovsNoB, fullCovsModelNoB,
165  fullCovWeightsNoB, gainDict, noiseDict, aDict, bDict, pdfPages)
166  self.plotNormalizedCovariancesplotNormalizedCovariances(0, 0, mu, fullCovs, fullCovsModel, fullCovWeights, fullCovsNoB,
167  fullCovsModelNoB, fullCovWeightsNoB, pdfPages,
168  offset=0.01, topPlot=True,
169  numberOfBins=self.plotNormalizedCovariancesNumberOfBinsplotNormalizedCovariancesNumberOfBins,
170  log=log)
171  self.plotNormalizedCovariancesplotNormalizedCovariances(0, 1, mu, fullCovs, fullCovsModel, fullCovWeights, fullCovsNoB,
172  fullCovsModelNoB, fullCovWeightsNoB, pdfPages,
173  numberOfBins=self.plotNormalizedCovariancesNumberOfBinsplotNormalizedCovariancesNumberOfBins,
174  log=log)
175  self.plotNormalizedCovariancesplotNormalizedCovariances(1, 0, mu, fullCovs, fullCovsModel, fullCovWeights, fullCovsNoB,
176  fullCovsModelNoB, fullCovWeightsNoB, pdfPages,
177  numberOfBins=self.plotNormalizedCovariancesNumberOfBinsplotNormalizedCovariancesNumberOfBins,
178  log=log)
179  self.plot_a_bplot_a_b(aDict, bDict, pdfPages)
180  self.ab_vs_distab_vs_dist(aDict, bDict, pdfPages, bRange=4)
181  self.plotAcoeffsSumplotAcoeffsSum(aDict, bDict, pdfPages)
182  self.plotRelativeBiasACoeffsplotRelativeBiasACoeffs(aDict, aDictNoB, fullCovsModel, fullCovsModelNoB,
183  self.signalElectronsRelativeAsignalElectronsRelativeA, gainDict, pdfPages, maxr=4)
184 
185  return
186 
187  @staticmethod
188  def plotCovariances(mu, covs, covsModel, covsWeights, covsNoB, covsModelNoB, covsWeightsNoB,
189  gainDict, noiseDict, aDict, bDict, pdfPages):
190  """Plot covariances and models: Cov00, Cov10, Cov01.
191 
192  Figs. 6 and 7 of Astier+19
193 
194  Parameters
195  ----------
196  mu : `dict`, [`str`, `list`]
197  Dictionary keyed by amp name with mean signal values.
198 
199  covs : `dict`, [`str`, `list`]
200  Dictionary keyed by amp names containing a list of measued covariances per mean flux.
201 
202  covsModel : `dict`, [`str`, `list`]
203  Dictionary keyed by amp names containinging covariances model (Eq. 20 of Astier+19) per mean flux.
204 
205  covsWeights : `dict`, [`str`, `list`]
206  Dictionary keyed by amp names containinging sqrt. of covariances weights.
207 
208  covsNoB : `dict`, [`str`, `list`]
209  Dictionary keyed by amp names containing a list of measued covariances per mean flux ('b'=0 in
210  Astier+19).
211 
212  covsModelNoB : `dict`, [`str`, `list`]
213  Dictionary keyed by amp names containing covariances model (with 'b'=0 in Eq. 20 of Astier+19)
214  per mean flux.
215 
216  covsWeightsNoB : `dict`, [`str`, `list`]
217  Dictionary keyed by amp names containing sqrt. of covariances weights ('b' = 0 in Eq. 20 of
218  Astier+19).
219 
220  gainDict : `dict`, [`str`, `float`]
221  Dictionary keyed by amp names containing the gains in e-/ADU.
222 
223  noiseDict : `dict`, [`str`, `float`]
224  Dictionary keyed by amp names containing the rms redout noise in e-.
225 
226  aDict : `dict`, [`str`, `numpy.array`]
227  Dictionary keyed by amp names containing 'a' coefficients (Eq. 20 of Astier+19).
228 
229  bDict : `dict`, [`str`, `numpy.array`]
230  Dictionary keyed by amp names containing 'b' coefficients (Eq. 20 of Astier+19).
231 
232  pdfPages: `matplotlib.backends.backend_pdf.PdfPages`
233  PDF file where the plots will be saved.
234  """
235 
236  legendFontSize = 6.5
237  labelFontSize = 7
238  titleFontSize = 9
239  supTitleFontSize = 18
240  markerSize = 25
241 
242  nAmps = len(covs)
243  if nAmps == 2:
244  nRows, nCols = 2, 1
245  nRows = np.sqrt(nAmps)
246  mantissa, _ = np.modf(nRows)
247  if mantissa > 0:
248  nRows = int(nRows) + 1
249  nCols = nRows
250  else:
251  nRows = int(nRows)
252  nCols = nRows
253 
254  f, ax = plt.subplots(nrows=nRows, ncols=nCols, sharex='col', sharey='row', figsize=(13, 10))
255  f2, ax2 = plt.subplots(nrows=nRows, ncols=nCols, sharex='col', sharey='row', figsize=(13, 10))
256  fResCov00, axResCov00 = plt.subplots(nrows=nRows, ncols=nCols, sharex='col', sharey='row',
257  figsize=(13, 10))
258  fCov01, axCov01 = plt.subplots(nrows=nRows, ncols=nCols, sharex='col', sharey='row', figsize=(13, 10))
259  fCov10, axCov10 = plt.subplots(nrows=nRows, ncols=nCols, sharex='col', sharey='row', figsize=(13, 10))
260 
261  assert(len(covsModel) == nAmps)
262  assert(len(covsWeights) == nAmps)
263 
264  assert(len(covsNoB) == nAmps)
265  assert(len(covsModelNoB) == nAmps)
266  assert(len(covsWeightsNoB) == nAmps)
267 
268  for i, (amp, a, a2, aResVar, a3, a4) in enumerate(zip(covs, ax.flatten(),
269  ax2.flatten(), axResCov00.flatten(),
270  axCov01.flatten(), axCov10.flatten())):
271 
272  muAmp, cov, model, weight = mu[amp], covs[amp], covsModel[amp], covsWeights[amp]
273  if not np.isnan(np.array(cov)).all(): # If all the entries are np.nan, this is a bad amp.
274  aCoeffs, bCoeffs = np.array(aDict[amp]), np.array(bDict[amp])
275  gain, noise = gainDict[amp], noiseDict[amp]
276  (meanVecFinal, varVecFinal, varVecModelFinal,
277  varWeightsFinal, _) = getFitDataFromCovariances(0, 0, muAmp, cov, model, weight,
278  returnMasked=True)
279 
280  # Get weighted reduced chi2
281  chi2FullModelVar = calculateWeightedReducedChi2(varVecFinal, varVecModelFinal,
282  varWeightsFinal, len(meanVecFinal), 4)
283 
284  (meanVecFinalCov01, varVecFinalCov01, varVecModelFinalCov01,
285  _, _) = getFitDataFromCovariances(0, 0, muAmp, cov, model, weight, returnMasked=True)
286 
287  (meanVecFinalCov10, varVecFinalCov10, varVecModelFinalCov10,
288  _, _) = getFitDataFromCovariances(1, 0, muAmp, cov, model, weight, returnMasked=True)
289 
290  # cuadratic fit for residuals below
291  par2 = np.polyfit(meanVecFinal, varVecFinal, 2, w=varWeightsFinal)
292  varModelFinalQuadratic = np.polyval(par2, meanVecFinal)
293  chi2QuadModelVar = calculateWeightedReducedChi2(varVecFinal, varModelFinalQuadratic,
294  varWeightsFinal, len(meanVecFinal), 3)
295 
296  # fit with no 'b' coefficient (c = a*b in Eq. 20 of Astier+19)
297  covNoB, modelNoB, weightNoB = covsNoB[amp], covsModelNoB[amp], covsWeightsNoB[amp]
298  (meanVecFinalNoB, varVecFinalNoB, varVecModelFinalNoB,
299  varWeightsFinalNoB, _) = getFitDataFromCovariances(0, 0, muAmp, covNoB, modelNoB,
300  weightNoB, returnMasked=True)
301 
302  chi2FullModelNoBVar = calculateWeightedReducedChi2(varVecFinalNoB, varVecModelFinalNoB,
303  varWeightsFinalNoB, len(meanVecFinalNoB),
304  3)
305  stringLegend = (f"Gain: {gain:.4} e/ADU \n"
306  f"Noise: {noise:.4} e \n"
307  + r"$a_{00}$: %.3e 1/e"%aCoeffs[0, 0] + "\n"
308  + r"$b_{00}$: %.3e 1/e"%bCoeffs[0, 0]
309  + f"\nLast in fit: {meanVecFinal[-1]:.7} ADU ")
310  minMeanVecFinal = np.nanmin(meanVecFinal)
311  maxMeanVecFinal = np.nanmax(meanVecFinal)
312  deltaXlim = maxMeanVecFinal - minMeanVecFinal
313 
314  a.set_xlabel(r'Mean signal ($\mu$, ADU)', fontsize=labelFontSize)
315  a.set_ylabel(r'Variance (ADU$^2$)', fontsize=labelFontSize)
316  a.tick_params(labelsize=11)
317  a.set_xscale('linear')
318  a.set_yscale('linear')
319  a.scatter(meanVecFinal, varVecFinal, c='blue', marker='o', s=markerSize)
320  a.plot(meanVecFinal, varVecModelFinal, color='red', lineStyle='-')
321  a.text(0.03, 0.7, stringLegend, transform=a.transAxes, fontsize=legendFontSize)
322  a.set_title(amp, fontsize=titleFontSize)
323  a.set_xlim([minMeanVecFinal - 0.2*deltaXlim, maxMeanVecFinal + 0.2*deltaXlim])
324 
325  # Same as above, but in log-scale
326  a2.set_xlabel(r'Mean Signal ($\mu$, ADU)', fontsize=labelFontSize)
327  a2.set_ylabel(r'Variance (ADU$^2$)', fontsize=labelFontSize)
328  a2.tick_params(labelsize=11)
329  a2.set_xscale('log')
330  a2.set_yscale('log')
331  a2.plot(meanVecFinal, varVecModelFinal, color='red', lineStyle='-')
332  a2.scatter(meanVecFinal, varVecFinal, c='blue', marker='o', s=markerSize)
333  a2.text(0.03, 0.7, stringLegend, transform=a2.transAxes, fontsize=legendFontSize)
334  a2.set_title(amp, fontsize=titleFontSize)
335  a2.set_xlim([minMeanVecFinal, maxMeanVecFinal])
336 
337  # Residuals var - model
338  aResVar.set_xlabel(r'Mean signal ($\mu$, ADU)', fontsize=labelFontSize)
339  aResVar.set_ylabel(r'Residuals (ADU$^2$)', fontsize=labelFontSize)
340  aResVar.tick_params(labelsize=11)
341  aResVar.set_xscale('linear')
342  aResVar.set_yscale('linear')
343  aResVar.plot(meanVecFinal, varVecFinal - varVecModelFinal, color='blue', lineStyle='-',
344  label=r'Full fit ($\chi_{\rm{red}}^2$: %g)'%chi2FullModelVar)
345  aResVar.plot(meanVecFinal, varVecFinal - varModelFinalQuadratic, color='red', lineStyle='-',
346  label=r'Quadratic fit ($\chi_{\rm{red}}^2$: %g)'%chi2QuadModelVar)
347  aResVar.plot(meanVecFinalNoB, varVecFinalNoB - varVecModelFinalNoB, color='green',
348  lineStyle='-',
349  label=r'Full fit (b=0) ($\chi_{\rm{red}}^2$: %g)'%chi2FullModelNoBVar)
350  aResVar.axhline(color='black')
351  aResVar.set_title(amp, fontsize=titleFontSize)
352  aResVar.set_xlim([minMeanVecFinal - 0.2*deltaXlim, maxMeanVecFinal + 0.2*deltaXlim])
353  aResVar.legend(fontsize=7)
354 
355  a3.set_xlabel(r'Mean signal ($\mu$, ADU)', fontsize=labelFontSize)
356  a3.set_ylabel(r'Cov01 (ADU$^2$)', fontsize=labelFontSize)
357  a3.tick_params(labelsize=11)
358  a3.set_xscale('linear')
359  a3.set_yscale('linear')
360  a3.scatter(meanVecFinalCov01, varVecFinalCov01, c='blue', marker='o', s=markerSize)
361  a3.plot(meanVecFinalCov01, varVecModelFinalCov01, color='red', lineStyle='-')
362  a3.set_title(amp, fontsize=titleFontSize)
363  a3.set_xlim([minMeanVecFinal - 0.2*deltaXlim, maxMeanVecFinal + 0.2*deltaXlim])
364 
365  a4.set_xlabel(r'Mean signal ($\mu$, ADU)', fontsize=labelFontSize)
366  a4.set_ylabel(r'Cov10 (ADU$^2$)', fontsize=labelFontSize)
367  a4.tick_params(labelsize=11)
368  a4.set_xscale('linear')
369  a4.set_yscale('linear')
370  a4.scatter(meanVecFinalCov10, varVecFinalCov10, c='blue', marker='o', s=markerSize)
371  a4.plot(meanVecFinalCov10, varVecModelFinalCov10, color='red', lineStyle='-')
372  a4.set_title(amp, fontsize=titleFontSize)
373  a4.set_xlim([minMeanVecFinal - 0.2*deltaXlim, maxMeanVecFinal + 0.2*deltaXlim])
374 
375  else:
376  a.set_title(f"{amp} (BAD)", fontsize=titleFontSize)
377  a2.set_title(f"{amp} (BAD)", fontsize=titleFontSize)
378  a3.set_title(f"{amp} (BAD)", fontsize=titleFontSize)
379  a4.set_title(f"{amp} (BAD)", fontsize=titleFontSize)
380 
381  f.suptitle("PTC from covariances as in Astier+19 \n Fit: Eq. 20, Astier+19",
382  fontsize=supTitleFontSize)
383  pdfPages.savefig(f)
384  f2.suptitle("PTC from covariances as in Astier+19 (log-log) \n Fit: Eq. 20, Astier+19",
385  fontsize=supTitleFontSize)
386  pdfPages.savefig(f2)
387  fResCov00.suptitle("Residuals (data-model) for Cov00 (Var)", fontsize=supTitleFontSize)
388  pdfPages.savefig(fResCov00)
389  fCov01.suptitle("Cov01 as in Astier+19 (nearest parallel neighbor covariance) \n"
390  " Fit: Eq. 20, Astier+19", fontsize=supTitleFontSize)
391  pdfPages.savefig(fCov01)
392  fCov10.suptitle("Cov10 as in Astier+19 (nearest serial neighbor covariance) \n"
393  "Fit: Eq. 20, Astier+19", fontsize=supTitleFontSize)
394  pdfPages.savefig(fCov10)
395 
396  return
397 
398  def plotNormalizedCovariances(self, i, j, inputMu, covs, covsModel, covsWeights, covsNoB, covsModelNoB,
399  covsWeightsNoB, pdfPages, offset=0.004,
400  numberOfBins=10, plotData=True, topPlot=False, log=None):
401  """Plot C_ij/mu vs mu.
402 
403  Figs. 8, 10, and 11 of Astier+19
404 
405  Parameters
406  ----------
407  i : `int`
408  Covariane lag
409 
410  j : `int`
411  Covariance lag
412 
413  inputMu : `dict`, [`str`, `list`]
414  Dictionary keyed by amp name with mean signal values.
415 
416  covs : `dict`, [`str`, `list`]
417  Dictionary keyed by amp names containing a list of measued covariances per mean flux.
418 
419  covsModel : `dict`, [`str`, `list`]
420  Dictionary keyed by amp names containinging covariances model (Eq. 20 of Astier+19) per mean flux.
421 
422  covsWeights : `dict`, [`str`, `list`]
423  Dictionary keyed by amp names containinging sqrt. of covariances weights.
424 
425  covsNoB : `dict`, [`str`, `list`]
426  Dictionary keyed by amp names containing a list of measued covariances per mean flux ('b'=0 in
427  Astier+19).
428 
429  covsModelNoB : `dict`, [`str`, `list`]
430  Dictionary keyed by amp names containing covariances model (with 'b'=0 in Eq. 20 of Astier+19)
431  per mean flux.
432 
433  covsWeightsNoB : `dict`, [`str`, `list`]
434  Dictionary keyed by amp names containing sqrt. of covariances weights ('b' = 0 in Eq. 20 of
435  Astier+19).
436 
437  expIdMask : `dict`, [`str`, `list`]
438  Dictionary keyed by amp names containing the masked exposure pairs.
439 
440  pdfPages: `matplotlib.backends.backend_pdf.PdfPages`
441  PDF file where the plots will be saved.
442 
443  offset : `float`, optional
444  Constant offset factor to plot covariances in same panel (so they don't overlap).
445 
446  numberOfBins : `int`, optional
447  Number of bins for top and bottom plot.
448 
449  plotData : `bool`, optional
450  Plot the data points?
451 
452  topPlot : `bool`, optional
453  Plot the top plot with the covariances, and the bottom plot with the model residuals?
454 
455  log : `lsst.log.Log`, optional
456  Logger to handle messages.
457  """
458  if not topPlot:
459  fig = plt.figure(figsize=(8, 10))
460  gs = gridspec.GridSpec(2, 1, height_ratios=[3, 1])
461  gs.update(hspace=0)
462  ax0 = plt.subplot(gs[0])
463  plt.setp(ax0.get_xticklabels(), visible=False)
464  else:
465  fig = plt.figure(figsize=(8, 8))
466  ax0 = plt.subplot(111)
467  ax0.ticklabel_format(style='sci', axis='x', scilimits=(0, 0))
468  ax0.tick_params(axis='both', labelsize='x-large')
469  mue, rese, wce = [], [], []
470  mueNoB, reseNoB, wceNoB = [], [], []
471  for counter, amp in enumerate(covs):
472  muAmp, fullCov, fullCovModel, fullCovWeight = (inputMu[amp], covs[amp], covsModel[amp],
473  covsWeights[amp])
474  if len(fullCov) == 0:
475  continue
476  mu, cov, model, weightCov, _ = getFitDataFromCovariances(i, j, muAmp, fullCov, fullCovModel,
477  fullCovWeight, divideByMu=True,
478  returnMasked=True)
479 
480  mue += list(mu)
481  rese += list(cov - model)
482  wce += list(weightCov)
483 
484  fullCovNoB, fullCovModelNoB, fullCovWeightNoB = (covsNoB[amp], covsModelNoB[amp],
485  covsWeightsNoB[amp])
486  if len(fullCovNoB) == 0:
487  continue
488  (muNoB, covNoB, modelNoB,
489  weightCovNoB, _) = getFitDataFromCovariances(i, j, muAmp, fullCovNoB, fullCovModelNoB,
490  fullCovWeightNoB, divideByMu=True,
491  returnMasked=True)
492 
493  mueNoB += list(muNoB)
494  reseNoB += list(covNoB - modelNoB)
495  wceNoB += list(weightCovNoB)
496 
497  # the corresponding fit
498  fit_curve, = plt.plot(mu, model + counter*offset, '-', linewidth=4.0)
499  # bin plot. len(mu) = no binning
500  gind = self.indexForBinsindexForBins(mu, numberOfBins)
501 
502  xb, yb, wyb, sigyb = self.binDatabinData(mu, cov, gind, weightCov)
503  plt.errorbar(xb, yb+counter*offset, yerr=sigyb, marker='o', linestyle='none', markersize=6.5,
504  color=fit_curve.get_color(), label=f"{amp} (N: {len(mu)})")
505  # plot the data
506  if plotData:
507  points, = plt.plot(mu, cov + counter*offset, '.', color=fit_curve.get_color())
508  plt.legend(loc='upper right', fontsize=8)
509  # end loop on amps
510  mue = np.array(mue)
511  rese = np.array(rese)
512  wce = np.array(wce)
513  mueNoB = np.array(mueNoB)
514  reseNoB = np.array(reseNoB)
515  wceNoB = np.array(wceNoB)
516 
517  plt.xlabel(r"$\mu (el)$", fontsize='x-large')
518  plt.ylabel(r"$Cov{%d%d}/\mu + Cst (el)$"%(i, j), fontsize='x-large')
519  if (not topPlot):
520  gind = self.indexForBinsindexForBins(mue, numberOfBins)
521  xb, yb, wyb, sigyb = self.binDatabinData(mue, rese, gind, wce)
522 
523  ax1 = plt.subplot(gs[1], sharex=ax0)
524  ax1.errorbar(xb, yb, yerr=sigyb, marker='o', linestyle='none', label='Full fit')
525  gindNoB = self.indexForBinsindexForBins(mueNoB, numberOfBins)
526  xb2, yb2, wyb2, sigyb2 = self.binDatabinData(mueNoB, reseNoB, gindNoB, wceNoB)
527 
528  ax1.errorbar(xb2, yb2, yerr=sigyb2, marker='o', linestyle='none', label='b = 0')
529  ax1.tick_params(axis='both', labelsize='x-large')
530  plt.legend(loc='upper left', fontsize='large')
531  # horizontal line at zero
532  plt.plot(xb, [0]*len(xb), '--', color='k')
533  plt.ticklabel_format(style='sci', axis='x', scilimits=(0, 0))
534  plt.ticklabel_format(style='sci', axis='y', scilimits=(0, 0))
535  plt.xlabel(r'$\mu (el)$', fontsize='x-large')
536  plt.ylabel(r'$Cov{%d%d}/\mu$ -model (el)'%(i, j), fontsize='x-large')
537  plt.tight_layout()
538  plt.suptitle(f"Nbins: {numberOfBins}")
539  # overlapping y labels:
540  fig.canvas.draw()
541  labels0 = [item.get_text() for item in ax0.get_yticklabels()]
542  labels0[0] = u''
543  ax0.set_yticklabels(labels0)
544  pdfPages.savefig(fig)
545 
546  return
547 
548  @staticmethod
549  def plot_a_b(aDict, bDict, pdfPages, bRange=3):
550  """Fig. 12 of Astier+19
551 
552  Color display of a and b arrays fits, averaged over channels.
553 
554  Parameters
555  ----------
556  aDict : `dict`, [`numpy.array`]
557  Dictionary keyed by amp names containing the fitted 'a' coefficients from the model
558  in Eq. 20 of Astier+19 (if `ptcFitType` is `FULLCOVARIANCE`).
559 
560  bDict : `dict`, [`numpy.array`]
561  Dictionary keyed by amp names containing the fitted 'b' coefficients from the model
562  in Eq. 20 of Astier+19 (if `ptcFitType` is `FULLCOVARIANCE`).
563 
564  pdfPages: `matplotlib.backends.backend_pdf.PdfPages`
565  PDF file where the plots will be saved.
566 
567  bRange : `int`
568  Maximum lag for b arrays.
569  """
570  a, b = [], []
571  for amp in aDict:
572  if np.isnan(aDict[amp]).all():
573  continue
574  a.append(aDict[amp])
575  b.append(bDict[amp])
576  a = np.array(a).mean(axis=0)
577  b = np.array(b).mean(axis=0)
578  fig = plt.figure(figsize=(7, 11))
579  ax0 = fig.add_subplot(2, 1, 1)
580  im0 = ax0.imshow(np.abs(a.transpose()), origin='lower', norm=mpl.colors.LogNorm())
581  ax0.tick_params(axis='both', labelsize='x-large')
582  ax0.set_title(r'$|a|$', fontsize='x-large')
583  ax0.xaxis.set_ticks_position('bottom')
584  cb0 = plt.colorbar(im0)
585  cb0.ax.tick_params(labelsize='x-large')
586 
587  ax1 = fig.add_subplot(2, 1, 2)
588  ax1.tick_params(axis='both', labelsize='x-large')
589  ax1.yaxis.set_major_locator(MaxNLocator(integer=True))
590  ax1.xaxis.set_major_locator(MaxNLocator(integer=True))
591  im1 = ax1.imshow(1e6*b[:bRange, :bRange].transpose(), origin='lower')
592  cb1 = plt.colorbar(im1)
593  cb1.ax.tick_params(labelsize='x-large')
594  ax1.set_title(r'$b \times 10^6$', fontsize='x-large')
595  ax1.xaxis.set_ticks_position('bottom')
596  plt.tight_layout()
597  pdfPages.savefig(fig)
598 
599  return
600 
601  @staticmethod
602  def ab_vs_dist(aDict, bDict, pdfPages, bRange=4):
603  """Fig. 13 of Astier+19.
604 
605  Values of a and b arrays fits, averaged over amplifiers, as a function of distance.
606 
607  Parameters
608  ----------
609  aDict : `dict`, [`numpy.array`]
610  Dictionary keyed by amp names containing the fitted 'a' coefficients from the model
611  in Eq. 20 of Astier+19 (if `ptcFitType` is `FULLCOVARIANCE`).
612 
613  bDict : `dict`, [`numpy.array`]
614  Dictionary keyed by amp names containing the fitted 'b' coefficients from the model
615  in Eq. 20 of Astier+19 (if `ptcFitType` is `FULLCOVARIANCE`).
616 
617  pdfPages: `matplotlib.backends.backend_pdf.PdfPages`
618  PDF file where the plots will be saved.
619 
620  bRange : `int`
621  Maximum lag for b arrays.
622  """
623  assert (len(aDict) == len(bDict))
624  a = []
625  for amp in aDict:
626  if np.isnan(aDict[amp]).all():
627  continue
628  a.append(aDict[amp])
629  a = np.array(a)
630  y = a.mean(axis=0)
631  sy = a.std(axis=0)/np.sqrt(len(aDict))
632  i, j = np.indices(y.shape)
633  upper = (i >= j).ravel()
634  r = np.sqrt(i**2 + j**2).ravel()
635  y = y.ravel()
636  sy = sy.ravel()
637  fig = plt.figure(figsize=(6, 9))
638  ax = fig.add_subplot(211)
639  ax.set_xlim([0.5, r.max()+1])
640  ax.errorbar(r[upper], y[upper], yerr=sy[upper], marker='o', linestyle='none', color='b',
641  label='$i>=j$')
642  ax.errorbar(r[~upper], y[~upper], yerr=sy[~upper], marker='o', linestyle='none', color='r',
643  label='$i<j$')
644  ax.legend(loc='upper center', fontsize='x-large')
645  ax.set_xlabel(r'$\sqrt{i^2+j^2}$', fontsize='x-large')
646  ax.set_ylabel(r'$a_{ij}$', fontsize='x-large')
647  ax.set_yscale('log')
648  ax.tick_params(axis='both', labelsize='x-large')
649 
650  #
651  axb = fig.add_subplot(212)
652  b = []
653  for amp in bDict:
654  if np.isnan(bDict[amp]).all():
655  continue
656  b.append(bDict[amp])
657  b = np.array(b)
658  yb = b.mean(axis=0)
659  syb = b.std(axis=0)/np.sqrt(len(bDict))
660  ib, jb = np.indices(yb.shape)
661  upper = (ib > jb).ravel()
662  rb = np.sqrt(i**2 + j**2).ravel()
663  yb = yb.ravel()
664  syb = syb.ravel()
665  xmin = -0.2
666  xmax = bRange
667  axb.set_xlim([xmin, xmax+0.2])
668  cutu = (r > xmin) & (r < xmax) & (upper)
669  cutl = (r > xmin) & (r < xmax) & (~upper)
670  axb.errorbar(rb[cutu], yb[cutu], yerr=syb[cutu], marker='o', linestyle='none', color='b',
671  label='$i>=j$')
672  axb.errorbar(rb[cutl], yb[cutl], yerr=syb[cutl], marker='o', linestyle='none', color='r',
673  label='$i<j$')
674  plt.legend(loc='upper center', fontsize='x-large')
675  axb.set_xlabel(r'$\sqrt{i^2+j^2}$', fontsize='x-large')
676  axb.set_ylabel(r'$b_{ij}$', fontsize='x-large')
677  axb.ticklabel_format(style='sci', axis='y', scilimits=(0, 0))
678  axb.tick_params(axis='both', labelsize='x-large')
679  plt.tight_layout()
680  pdfPages.savefig(fig)
681 
682  return
683 
684  @staticmethod
685  def plotAcoeffsSum(aDict, bDict, pdfPages):
686  """Fig. 14. of Astier+19
687 
688  Cumulative sum of a_ij as a function of maximum separation. This plot displays the average over
689  channels.
690 
691  Parameters
692  ----------
693  aDict : `dict`, [`numpy.array`]
694  Dictionary keyed by amp names containing the fitted 'a' coefficients from the model
695  in Eq. 20 of Astier+19 (if `ptcFitType` is `FULLCOVARIANCE`).
696 
697  bDict : `dict`, [`numpy.array`]
698  Dictionary keyed by amp names containing the fitted 'b' coefficients from the model
699  in Eq. 20 of Astier+19 (if `ptcFitType` is `FULLCOVARIANCE`).
700 
701  pdfPages: `matplotlib.backends.backend_pdf.PdfPages`
702  PDF file where the plots will be saved.
703  """
704  assert (len(aDict) == len(bDict))
705  a, b = [], []
706  for amp in aDict:
707  if np.isnan(aDict[amp]).all() or np.isnan(bDict[amp]).all():
708  continue
709  a.append(aDict[amp])
710  b.append(bDict[amp])
711  a = np.array(a).mean(axis=0)
712  b = np.array(b).mean(axis=0)
713  fig = plt.figure(figsize=(7, 6))
714  w = 4*np.ones_like(a)
715  w[0, 1:] = 2
716  w[1:, 0] = 2
717  w[0, 0] = 1
718  wa = w*a
719  indices = range(1, a.shape[0]+1)
720  sums = [wa[0:n, 0:n].sum() for n in indices]
721  ax = plt.subplot(111)
722  ax.plot(indices, sums/sums[0], 'o', color='b')
723  ax.set_yscale('log')
724  ax.set_xlim(indices[0]-0.5, indices[-1]+0.5)
725  ax.set_ylim(None, 1.2)
726  ax.set_ylabel(r'$[\sum_{|i|<n\ &\ |j|<n} a_{ij}] / |a_{00}|$', fontsize='x-large')
727  ax.set_xlabel('n', fontsize='x-large')
728  ax.tick_params(axis='both', labelsize='x-large')
729  plt.tight_layout()
730  pdfPages.savefig(fig)
731 
732  return
733 
734  @staticmethod
735  def plotRelativeBiasACoeffs(aDict, aDictNoB, fullCovsModel, fullCovsModelNoB, signalElectrons,
736  gainDict, pdfPages, maxr=None):
737  """Fig. 15 in Astier+19.
738 
739  Illustrates systematic bias from estimating 'a'
740  coefficients from the slope of correlations as opposed to the
741  full model in Astier+19.
742 
743  Parameters
744  ----------
745  aDict: `dict`
746  Dictionary of 'a' matrices (Eq. 20, Astier+19), with amp names as keys.
747 
748  aDictNoB: `dict`
749  Dictionary of 'a' matrices ('b'= 0 in Eq. 20, Astier+19), with amp names as keys.
750 
751  fullCovsModel : `dict`, [`str`, `list`]
752  Dictionary keyed by amp names containing covariances model per mean flux.
753 
754  fullCovsModelNoB : `dict`, [`str`, `list`]
755  Dictionary keyed by amp names containing covariances model (with 'b'=0 in Eq. 20 of
756  Astier+19) per mean flux.
757 
758  signalElectrons : `float`
759  Signal at which to evaluate the a_ij coefficients.
760 
761  pdfPages: `matplotlib.backends.backend_pdf.PdfPages`
762  PDF file where the plots will be saved.
763 
764  gainDict : `dict`, [`str`, `float`]
765  Dicgionary keyed by amp names with the gains in e-/ADU.
766 
767  maxr : `int`, optional
768  Maximum lag.
769  """
770 
771  fig = plt.figure(figsize=(7, 11))
772  title = [f"'a' relative bias at {signalElectrons} e", "'a' relative bias (b=0)"]
773  data = [(aDict, fullCovsModel), (aDictNoB, fullCovsModelNoB)]
774 
775  for k, pair in enumerate(data):
776  diffs = []
777  amean = []
778  for amp in pair[0]:
779  covModel = np.array(pair[1][amp])
780  if np.isnan(covModel).all():
781  continue
782  # Compute the "a" coefficients of the Antilogus+14 (1402.0725) model as in
783  # Guyonnet+15 (1501.01577, eq. 16, the slope of cov/var at a given flux mu in
784  # electrons). Eq. 16 of 1501.01577 is an approximation to the more complete
785  # model in Astier+19 (1905.08677).
786  var = covModel[0, 0, 0] # ADU^2
787  # For a result in electrons^-1, we have to use mu in electrons
788  aOld = covModel[0, :, :]/(var*signalElectrons)
789  a = pair[0][amp]
790  amean.append(a)
791  diffs.append((aOld-a))
792  amean = np.array(amean).mean(axis=0)
793  diff = np.array(diffs).mean(axis=0)
794  diff = diff/amean
795  diff = diff[:]
796  # The difference should be close to zero
797  diff[0, 0] = 0
798  if maxr is None:
799  maxr = diff.shape[0]
800  diff = diff[:maxr, :maxr]
801  ax0 = fig.add_subplot(2, 1, k+1)
802  im0 = ax0.imshow(diff.transpose(), origin='lower')
803  ax0.yaxis.set_major_locator(MaxNLocator(integer=True))
804  ax0.xaxis.set_major_locator(MaxNLocator(integer=True))
805  ax0.tick_params(axis='both', labelsize='x-large')
806  plt.colorbar(im0)
807  ax0.set_title(title[k])
808 
809  plt.tight_layout()
810  pdfPages.savefig(fig)
811 
812  return
813 
814  def _plotStandardPtc(self, dataset, ptcFitType, pdfPages):
815  """Plot PTC, var/signal vs signal, linearity, and linearity residual per amplifier.
816 
817  Parameters
818  ----------
819  dataset : `lsst.ip.isr.ptcDataset.PhotonTransferCurveDataset`
820  The dataset containing the means, variances, exposure times, and mask.
821 
822  ptcFitType : `str`
823  Type of the model fit to the PTC. Options: 'FULLCOVARIANCE', EXPAPPROXIMATION, or 'POLYNOMIAL'.
824 
825  pdfPages: `matplotlib.backends.backend_pdf.PdfPages`
826  PDF file where the plots will be saved.
827  """
828 
829  if ptcFitType == 'EXPAPPROXIMATION':
830  ptcFunc = funcAstier
831  stringTitle = (r"Var = $\frac{1}{2g^2a_{00}}(\exp (2a_{00} \mu g) - 1) + \frac{n_{00}}{g^2}$ ")
832  elif ptcFitType == 'POLYNOMIAL':
833  ptcFunc = funcPolynomial
834  for key in dataset.ptcFitPars:
835  deg = len(dataset.ptcFitPars[key]) - 1
836  break
837  stringTitle = r"Polynomial (degree: %g)" % (deg)
838  else:
839  raise RuntimeError(f"The input dataset had an invalid dataset.ptcFitType: {ptcFitType}. \n"
840  "Options: 'FULLCOVARIANCE', EXPAPPROXIMATION, or 'POLYNOMIAL'.")
841 
842  legendFontSize = 6.5
843  labelFontSize = 8
844  titleFontSize = 9
845  supTitleFontSize = 18
846  markerSize = 25
847 
848  # General determination of the size of the plot grid
849  nAmps = len(dataset.ampNames)
850  if nAmps == 2:
851  nRows, nCols = 2, 1
852  nRows = np.sqrt(nAmps)
853  mantissa, _ = np.modf(nRows)
854  if mantissa > 0:
855  nRows = int(nRows) + 1
856  nCols = nRows
857  else:
858  nRows = int(nRows)
859  nCols = nRows
860 
861  f, ax = plt.subplots(nrows=nRows, ncols=nCols, sharex='col', sharey='row', figsize=(13, 10))
862  f2, ax2 = plt.subplots(nrows=nRows, ncols=nCols, sharex='col', sharey='row', figsize=(13, 10))
863  f3, ax3 = plt.subplots(nrows=nRows, ncols=nCols, sharex='col', sharey='row', figsize=(13, 10))
864 
865  for i, (amp, a, a2, a3) in enumerate(zip(dataset.ampNames, ax.flatten(), ax2.flatten(),
866  ax3.flatten())):
867  meanVecOriginal = np.ravel(np.array(dataset.rawMeans[amp]))
868  varVecOriginal = np.ravel(np.array(dataset.rawVars[amp]))
869  mask = np.ravel(np.array(dataset.expIdMask[amp]))
870  if np.isnan(mask[0]): # All NaNs the whole amp is bad
871  a.set_title(f"{amp} (BAD)", fontsize=titleFontSize)
872  a2.set_title(f"{amp} (BAD)", fontsize=titleFontSize)
873  a3.set_title(f"{amp} (BAD)", fontsize=titleFontSize)
874  continue
875  else:
876  mask = mask.astype(bool)
877  meanVecFinal = meanVecOriginal[mask]
878  varVecFinal = varVecOriginal[mask]
879  meanVecOutliers = meanVecOriginal[np.invert(mask)]
880  varVecOutliers = varVecOriginal[np.invert(mask)]
881  pars, parsErr = np.array(dataset.ptcFitPars[amp]), np.array(dataset.ptcFitParsError[amp])
882  ptcRedChi2 = dataset.ptcFitChiSq[amp]
883  if ptcFitType == 'EXPAPPROXIMATION':
884  if len(meanVecFinal):
885  ptcA00, ptcA00error = pars[0], parsErr[0]
886  ptcGain, ptcGainError = pars[1], parsErr[1]
887  ptcNoise = np.sqrt((pars[2])) # pars[2] is in (e-)^2
888  ptcNoiseAdu = ptcNoise*(1./ptcGain)
889  ptcNoiseError = 0.5*(parsErr[2]/np.fabs(pars[2]))*np.sqrt(np.fabs(pars[2]))
890  stringLegend = (f"a00: {ptcA00:.2e}+/-{ptcA00error:.2e} 1/e"
891  f"\nGain: {ptcGain:.4}+/-{ptcGainError:.2e} e/ADU"
892  f"\nNoise: {ptcNoise:.4}+/-{ptcNoiseError:.2e} e\n"
893  r"$\chi^2_{\rm{red}}$: " + f"{ptcRedChi2:.4}"
894  f"\nLast in fit: {meanVecFinal[-1]:.7} ADU ")
895 
896  if ptcFitType == 'POLYNOMIAL':
897  if len(meanVecFinal):
898  ptcGain, ptcGainError = 1./pars[1], np.fabs(1./pars[1])*(parsErr[1]/pars[1])
899  ptcNoiseAdu = np.sqrt((pars[0])) # pars[0] is in ADU^2
900  ptcNoise = ptcNoiseAdu*ptcGain
901  ptcNoiseError = (0.5*(parsErr[0]/np.fabs(pars[0]))*(np.sqrt(np.fabs(pars[0]))))*ptcGain
902  stringLegend = (f"Gain: {ptcGain:.4}+/-{ptcGainError:.2e} e/ADU\n"
903  f"Noise: {ptcNoise:.4}+/-{ptcNoiseError:.2e} e\n"
904  r"$\chi^2_{\rm{red}}$: " + f"{ptcRedChi2:.4}"
905  f"\nLast in fit: {meanVecFinal[-1]:.7} ADU ")
906 
907  a.set_xlabel(r'Mean signal ($\mu$, ADU)', fontsize=labelFontSize)
908  a.set_ylabel(r'Variance (ADU$^2$)', fontsize=labelFontSize)
909  a.tick_params(labelsize=11)
910  a.set_xscale('linear')
911  a.set_yscale('linear')
912 
913  a2.set_xlabel(r'Mean Signal ($\mu$, ADU)', fontsize=labelFontSize)
914  a2.set_ylabel(r'Variance (ADU$^2$)', fontsize=labelFontSize)
915  a2.tick_params(labelsize=11)
916  a2.set_xscale('log')
917  a2.set_yscale('log')
918 
919  a3.set_xlabel(r'Mean signal ($\mu$, ADU)', fontsize=labelFontSize)
920  a3.set_ylabel(r'Variance/$\mu$ (ADU)', fontsize=labelFontSize)
921  a3.tick_params(labelsize=11)
922  a3.set_xscale('log')
923  a3.set_yscale('linear')
924 
925  minMeanVecFinal = np.nanmin(meanVecFinal)
926  maxMeanVecFinal = np.nanmax(meanVecFinal)
927  meanVecFit = np.linspace(minMeanVecFinal, maxMeanVecFinal, 100*len(meanVecFinal))
928  minMeanVecOriginal = np.nanmin(meanVecOriginal)
929  maxMeanVecOriginal = np.nanmax(meanVecOriginal)
930  deltaXlim = maxMeanVecOriginal - minMeanVecOriginal
931  a.plot(meanVecFit, ptcFunc(pars, meanVecFit), color='red')
932  a.plot(meanVecFinal, ptcNoiseAdu**2 + (1./ptcGain)*meanVecFinal, color='green',
933  linestyle='--')
934  a.scatter(meanVecFinal, varVecFinal, c='blue', marker='o', s=markerSize)
935  a.scatter(meanVecOutliers, varVecOutliers, c='magenta', marker='s', s=markerSize)
936  a.text(0.03, 0.66, stringLegend, transform=a.transAxes, fontsize=legendFontSize)
937  a.set_title(amp, fontsize=titleFontSize)
938  a.set_xlim([minMeanVecOriginal - 0.2*deltaXlim, maxMeanVecOriginal + 0.2*deltaXlim])
939 
940  # Same, but in log-scale
941  a2.plot(meanVecFit, ptcFunc(pars, meanVecFit), color='red')
942  a2.scatter(meanVecFinal, varVecFinal, c='blue', marker='o', s=markerSize)
943  a2.scatter(meanVecOutliers, varVecOutliers, c='magenta', marker='s', s=markerSize)
944  a2.text(0.03, 0.66, stringLegend, transform=a2.transAxes, fontsize=legendFontSize)
945  a2.set_title(amp, fontsize=titleFontSize)
946  a2.set_xlim([minMeanVecOriginal, maxMeanVecOriginal])
947 
948  # Var/mu vs mu
949  a3.plot(meanVecFit, ptcFunc(pars, meanVecFit)/meanVecFit, color='red')
950  a3.scatter(meanVecFinal, varVecFinal/meanVecFinal, c='blue', marker='o', s=markerSize)
951  a3.scatter(meanVecOutliers, varVecOutliers/meanVecOutliers, c='magenta', marker='s',
952  s=markerSize)
953  a3.text(0.05, 0.1, stringLegend, transform=a3.transAxes, fontsize=legendFontSize)
954  a3.set_title(amp, fontsize=titleFontSize)
955  a3.set_xlim([minMeanVecOriginal - 0.2*deltaXlim, maxMeanVecOriginal + 0.2*deltaXlim])
956 
957  f.suptitle("PTC \n Fit: " + stringTitle, fontsize=supTitleFontSize)
958  pdfPages.savefig(f)
959  f2.suptitle("PTC (log-log)", fontsize=supTitleFontSize)
960  pdfPages.savefig(f2)
961  f3.suptitle(r"Var/$\mu$", fontsize=supTitleFontSize)
962  pdfPages.savefig(f3)
963 
964  return
965 
966  def _plotLinearizer(self, dataset, linearizer, pdfPages):
967  """Plot linearity and linearity residual per amplifier
968 
969  Parameters
970  ----------
971  dataset : `lsst.ip.isr.ptcDataset.PhotonTransferCurveDataset`
972  The dataset containing the means, variances, exposure times, and mask.
973 
974  linearizer : `lsst.ip.isr.Linearizer`
975  Linearizer object
976  """
977  legendFontSize = 7
978  labelFontSize = 7
979  titleFontSize = 9
980  supTitleFontSize = 18
981 
982  # General determination of the size of the plot grid
983  nAmps = len(dataset.ampNames)
984  if nAmps == 2:
985  nRows, nCols = 2, 1
986  nRows = np.sqrt(nAmps)
987  mantissa, _ = np.modf(nRows)
988  if mantissa > 0:
989  nRows = int(nRows) + 1
990  nCols = nRows
991  else:
992  nRows = int(nRows)
993  nCols = nRows
994 
995  # Plot mean vs time (f1), and fractional residuals (f2)
996  f, ax = plt.subplots(nrows=nRows, ncols=nCols, sharex='col', sharey='row', figsize=(13, 10))
997  f2, ax2 = plt.subplots(nrows=nRows, ncols=nCols, sharex='col', sharey='row', figsize=(13, 10))
998  for i, (amp, a, a2) in enumerate(zip(dataset.ampNames, ax.flatten(), ax2.flatten())):
999  mask = dataset.expIdMask[amp]
1000  if np.isnan(mask[0]):
1001  a.set_title(f"{amp} (BAD)", fontsize=titleFontSize)
1002  a2.set_title(f"{amp} (BAD)", fontsize=titleFontSize)
1003  continue
1004  else:
1005  mask = mask.astype(bool)
1006  meanVecFinal = np.array(dataset.rawMeans[amp])[mask]
1007  timeVecFinal = np.array(dataset.rawExpTimes[amp])[mask]
1008 
1009  a.set_xlabel('Time (sec)', fontsize=labelFontSize)
1010  a.set_ylabel(r'Mean signal ($\mu$, ADU)', fontsize=labelFontSize)
1011  a.tick_params(labelsize=labelFontSize)
1012  a.set_xscale('linear')
1013  a.set_yscale('linear')
1014 
1015  a2.axhline(y=0, color='k')
1016  a2.axvline(x=0, color='k', linestyle='-')
1017  a2.set_xlabel(r'Mean signal ($\mu$, ADU)', fontsize=labelFontSize)
1018  a2.set_ylabel('Fractional nonlinearity (%)', fontsize=labelFontSize)
1019  a2.tick_params(labelsize=labelFontSize)
1020  a2.set_xscale('linear')
1021  a2.set_yscale('linear')
1022 
1023  pars, parsErr = linearizer.fitParams[amp], linearizer.fitParamsErr[amp]
1024  k0, k0Error = pars[0], parsErr[0]
1025  k1, k1Error = pars[1], parsErr[1]
1026  k2, k2Error = pars[2], parsErr[2]
1027  linRedChi2 = linearizer.fitChiSq[amp]
1028  stringLegend = (f"k0: {k0:.4}+/-{k0Error:.2e} ADU\nk1: {k1:.4}+/-{k1Error:.2e} ADU/t"
1029  f"\nk2: {k2:.2e}+/-{k2Error:.2e} ADU/t^2\n"
1030  r"$\chi^2_{\rm{red}}$: " + f"{linRedChi2:.4}")
1031  a.scatter(timeVecFinal, meanVecFinal)
1032  a.plot(timeVecFinal, funcPolynomial(pars, timeVecFinal), color='red')
1033  a.text(0.03, 0.75, stringLegend, transform=a.transAxes, fontsize=legendFontSize)
1034  a.set_title(f"{amp}", fontsize=titleFontSize)
1035 
1036  linearPart = k0 + k1*timeVecFinal
1037  fracLinRes = 100*(linearPart - meanVecFinal)/linearPart
1038  a2.plot(meanVecFinal, fracLinRes, c='g')
1039  a2.set_title(f"{amp}", fontsize=titleFontSize)
1040 
1041  f.suptitle("Linearity \n Fit: Polynomial (degree: %g)"
1042  % (len(pars)-1),
1043  fontsize=supTitleFontSize)
1044  f2.suptitle(r"Fractional NL residual" "\n"
1045  r"$100\times \frac{(k_0 + k_1*Time-\mu)}{k_0+k_1*Time}$",
1046  fontsize=supTitleFontSize)
1047  pdfPages.savefig(f)
1048  pdfPages.savefig(f2)
1049 
1050  @staticmethod
1051  def findGroups(x, maxDiff):
1052  """Group data into bins, with at most maxDiff distance between bins.
1053 
1054  Parameters
1055  ----------
1056  x: `list`
1057  Data to bin.
1058 
1059  maxDiff: `int`
1060  Maximum distance between bins.
1061 
1062  Returns
1063  -------
1064  index: `list`
1065  Bin indices.
1066  """
1067  ix = np.argsort(x)
1068  xsort = np.sort(x)
1069  index = np.zeros_like(x, dtype=np.int32)
1070  xc = xsort[0]
1071  group = 0
1072  ng = 1
1073 
1074  for i in range(1, len(ix)):
1075  xval = xsort[i]
1076  if (xval - xc < maxDiff):
1077  xc = (ng*xc + xval)/(ng+1)
1078  ng += 1
1079  index[ix[i]] = group
1080  else:
1081  group += 1
1082  ng = 1
1083  index[ix[i]] = group
1084  xc = xval
1085 
1086  return index
1087 
1088  @staticmethod
1089  def indexForBins(x, nBins):
1090  """Builds an index with regular binning. The result can be fed into binData.
1091 
1092  Parameters
1093  ----------
1094  x: `numpy.array`
1095  Data to bin.
1096  nBins: `int`
1097  Number of bin.
1098 
1099  Returns
1100  -------
1101  np.digitize(x, bins): `numpy.array`
1102  Bin indices.
1103  """
1104 
1105  bins = np.linspace(x.min(), x.max() + abs(x.max() * 1e-7), nBins + 1)
1106  return np.digitize(x, bins)
1107 
1108  @staticmethod
1109  def binData(x, y, binIndex, wy=None):
1110  """Bin data (usually for display purposes).
1111 
1112  Patrameters
1113  -----------
1114  x: `numpy.array`
1115  Data to bin.
1116 
1117  y: `numpy.array`
1118  Data to bin.
1119 
1120  binIdex: `list`
1121  Bin number of each datum.
1122 
1123  wy: `numpy.array`
1124  Inverse rms of each datum to use when averaging (the actual weight is wy**2).
1125 
1126  Returns:
1127  -------
1128 
1129  xbin: `numpy.array`
1130  Binned data in x.
1131 
1132  ybin: `numpy.array`
1133  Binned data in y.
1134 
1135  wybin: `numpy.array`
1136  Binned weights in y, computed from wy's in each bin.
1137 
1138  sybin: `numpy.array`
1139  Uncertainty on the bin average, considering actual scatter, and ignoring weights.
1140  """
1141 
1142  if wy is None:
1143  wy = np.ones_like(x)
1144  binIndexSet = set(binIndex)
1145  w2 = wy*wy
1146  xw2 = x*(w2)
1147  xbin = np.array([xw2[binIndex == i].sum()/w2[binIndex == i].sum() for i in binIndexSet])
1148 
1149  yw2 = y*w2
1150  ybin = np.array([yw2[binIndex == i].sum()/w2[binIndex == i].sum() for i in binIndexSet])
1151 
1152  wybin = np.sqrt(np.array([w2[binIndex == i].sum() for i in binIndexSet]))
1153  sybin = np.array([y[binIndex == i].std()/np.sqrt(np.array([binIndex == i]).sum())
1154  for i in binIndexSet])
1155 
1156  return xbin, ybin, wybin, sybin
def _plotStandardPtc(self, dataset, ptcFitType, pdfPages)
Definition: plotPtc.py:814
def plotRelativeBiasACoeffs(aDict, aDictNoB, fullCovsModel, fullCovsModelNoB, signalElectrons, gainDict, pdfPages, maxr=None)
Definition: plotPtc.py:736
def __init__(self, datasetFilename, linearizerFileName=None, outDir='.', detNum=999, signalElectronsRelativeA=75000, plotNormalizedCovariancesNumberOfBins=10)
Definition: plotPtc.py:85
def plotCovariances(mu, covs, covsModel, covsWeights, covsNoB, covsModelNoB, covsWeightsNoB, gainDict, noiseDict, aDict, bDict, pdfPages)
Definition: plotPtc.py:189
def ab_vs_dist(aDict, bDict, pdfPages, bRange=4)
Definition: plotPtc.py:602
def plotAcoeffsSum(aDict, bDict, pdfPages)
Definition: plotPtc.py:685
def plotNormalizedCovariances(self, i, j, inputMu, covs, covsModel, covsWeights, covsNoB, covsModelNoB, covsWeightsNoB, pdfPages, offset=0.004, numberOfBins=10, plotData=True, topPlot=False, log=None)
Definition: plotPtc.py:400
def _plotLinearizer(self, dataset, linearizer, pdfPages)
Definition: plotPtc.py:966
def run(self, filenameFull, datasetPtc, linearizer=None, log=None)
Definition: plotPtc.py:116
def plot_a_b(aDict, bDict, pdfPages, bRange=3)
Definition: plotPtc.py:549
def covAstierMakeAllPlots(self, dataset, pdfPages, log=None)
Definition: plotPtc.py:133
daf::base::PropertyList * list
Definition: fits.cc:913
daf::base::PropertySet * set
Definition: fits.cc:912
def getFitDataFromCovariances(i, j, mu, fullCov, fullCovModel, fullCovSqrtWeights, gain=1.0, divideByMu=False, returnMasked=False)
def calculateWeightedReducedChi2(measured, model, weightsMeasured, nData, nParsModel)
Definition: utils.py:41
def funcPolynomial(pars, x)
Definition: utils.py:482
bool all(CoordinateExpr< N > const &expr) noexcept
Return true if all elements are true.
Definition: Log.h:706
Angle abs(Angle const &a)
Definition: Angle.h:106
STL namespace.