Coverage for python / lsst / analysis / tools / actions / vector / calcRhoStatistics.py: 38%
89 statements
« prev ^ index » next coverage.py v7.13.5, created at 2026-05-06 09:07 +0000
« prev ^ index » next coverage.py v7.13.5, created at 2026-05-06 09:07 +0000
1# This file is part of analysis_tools.
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/>.
22from __future__ import annotations
24__all__ = (
25 "CalcRhoStatistics",
26 "TreecorrConfig",
27)
29import logging
30from collections.abc import Mapping
31from typing import TYPE_CHECKING, Any, cast
33import numpy as np
34import treecorr # type: ignore[import]
35from deprecated.sphinx import deprecated
37from lsst.meas.algorithms.treecorrUtils import TreecorrConfig as TreecorrConfigNew
38from lsst.pex.config import ChoiceField, ConfigField, Field
40from ...interfaces import KeyedData, KeyedDataAction, Vector
41from .calcMomentSize import CalcMomentSize
42from .ellipticity import CalcE, CalcEDiff
43from .mathActions import FractionalDifference
45if TYPE_CHECKING:
46 from treecorr import GGCorrelation, KKCorrelation
48 from ...interfaces import KeyedDataSchema
50_LOG = logging.getLogger(__name__)
53# TO DO: Remove TreecorrConfig in here for next major release (DM-47072)
54@deprecated(
55 reason=(
56 "TreecorrConfig is no longer a part of analysis_tools (DM-45899). "
57 "Please use lsst.meas.algorithms.treecorrUtils.TreecorrConfig instead."
58 ),
59 version="v28.0",
60 category=FutureWarning,
61)
62class TreecorrConfig(TreecorrConfigNew):
63 pass
66class CalcRhoStatistics(KeyedDataAction):
67 r"""Calculate rho statistics.
69 Rho statistics refer to a collection of correlation functions involving
70 PSF ellipticity and size residuals. They quantify the contribution from PSF
71 leakage due to errors in PSF modeling to the weak lensing shear correlation
72 functions.
74 .. _rho_definitions:
76 The exact definitions of rho statistics as defined in [1]_ are given below.
78 .. math::
80 \rho_1(\theta) &= \left\langle
81 \delta e^*_{PSF}(x)
82 \delta e_{PSF}(x+\theta)
83 \right\rangle
85 \rho_2(\theta) &= \left\langle
86 e^*_{PSF}(x)
87 \delta e_{PSF}(x+\theta
88 \right\rangle
90 \rho_3(\theta) &= \left\langle
91 (e^*_{PSF}\frac{\delta T_{PSF}}{T_{PSF}}(x))
92 (e_{PSF}\frac{\delta T_{PSF}}{T_{PSF}})(x+\theta)
93 \right\rangle
95 \rho_4(\theta) &= \left\langle
96 \delta e^*_{PSF}(x)
97 (e_{PSF}\frac{\delta T_{PSF}}{T_{PSF}})(x+\theta)
98 \right\rangle
100 \rho_5(\theta) &= \left\langle
101 e^*_{PSF}(x)
102 (e_{PSF}\frac{\delta T_{PSF}}{T_{PSF}})(x+\theta)
103 \right\rangle
106 In addition to these five, we also compute the auto-correlation function of
107 the fractional size residuals and call it as the :math:`\rho'_3( \theta )`,
108 as referred to in Melchior et al. (2015) [2]_.
110 .. math::
112 \rho'_3(\theta) = \left\langle\frac{\delta T_{PSF}}{T_{PSF}}(x)
113 \frac{\delta T_{PSF}}{T_{PSF}}(x+\theta)
114 \right\rangle
117 The definition of ellipticity used in [1]_ correspond to shear-type,
118 which is typically smaller by a factor of 4 than using distortion-type.
120 References
121 ----------
123 .. [1] Jarvis, M., Sheldon, E., Zuntz, J., Kacprzak, T., Bridle, S. L.,
124 et. al (2016).
125 The DES Science Verification weak lensing shear catalogues
126 MNRAS, 460, 2245–2281.
127 https://doi.org/10.1093/mnras/stw990;
128 https://arxiv.org/abs/1507.05603
129 .. [2] Melchior, P., et. al (2015)
130 Mass and galaxy distributions of four massive galaxy clusters from
131 Dark Energy Survey Science Verification data
132 MNRAS, 449, no. 3, pp. 2219–2238.
133 https://doi:10.1093/mnras/stv398
134 https://arxiv.org/abs/1405.4285
135 """
137 colRa = Field[str](doc="RA column", default="coord_ra")
139 colDec = Field[str](doc="Dec column", default="coord_dec")
141 colXx = Field[str](doc="The column name to get the xx shape component from.", default="{band}_ixx")
143 colYy = Field[str](doc="The column name to get the yy shape component from.", default="{band}_iyy")
145 colXy = Field[str](doc="The column name to get the xy shape component from.", default="{band}_ixy")
147 colPsfXx = Field[str](
148 doc="The column name to get the PSF xx shape component from.", default="{band}_ixxPSF"
149 )
151 colPsfYy = Field[str](
152 doc="The column name to get the PSF yy shape component from.", default="{band}_iyyPSF"
153 )
155 colPsfXy = Field[str](
156 doc="The column name to get the PSF xy shape component from.", default="{band}_ixyPSF"
157 )
159 ellipticityType = ChoiceField[str](
160 doc="The type of ellipticity to calculate",
161 optional=False,
162 allowed={
163 "distortion": r"Distortion, measured as :math:`(I_{xx}-I_{yy})/(I_{xx}+I_{yy})`",
164 "shear": (
165 r"Shear, measured as :math:`(I_{xx}-I_{yy})/(I_{xx}+I_{yy}+2\sqrt{I_{xx}I_{yy}-I_{xy}^2})`"
166 ),
167 },
168 default="distortion",
169 )
171 sizeType = ChoiceField[str](
172 doc="The type of size to calculate",
173 default="trace",
174 allowed={
175 "trace": "trace radius",
176 "determinant": "determinant radius",
177 },
178 )
180 treecorr = ConfigField[TreecorrConfigNew](
181 doc="TreeCorr configuration",
182 )
184 def setDefaults(self):
185 super().setDefaults()
186 self.treecorr = TreecorrConfigNew()
187 self.treecorr.sep_units = "arcmin"
188 self.treecorr.max_sep = 100.0
190 def getInputSchema(self) -> KeyedDataSchema:
191 return (
192 (self.colRa, Vector),
193 (self.colDec, Vector),
194 (self.colXx, Vector),
195 (self.colYy, Vector),
196 (self.colXy, Vector),
197 (self.colPsfXx, Vector),
198 (self.colPsfYy, Vector),
199 (self.colPsfXy, Vector),
200 )
202 def __call__(self, data: KeyedData, **kwargs) -> KeyedData:
203 calcEMeas = CalcE(
204 colXx=self.colXx,
205 colYy=self.colYy,
206 colXy=self.colXy,
207 ellipticityType=self.ellipticityType,
208 )
209 calcEpsf = CalcE(
210 colXx=self.colPsfXx,
211 colYy=self.colPsfYy,
212 colXy=self.colPsfXy,
213 ellipticityType=self.ellipticityType,
214 )
216 calcEDiff = CalcEDiff(colA=calcEMeas, colB=calcEpsf)
218 calcSizeResidual = FractionalDifference(
219 actionA=CalcMomentSize(
220 colXx=self.colPsfXx,
221 colYy=self.colPsfYy,
222 colXy=self.colPsfXy,
223 sizeType=self.sizeType,
224 ),
225 actionB=CalcMomentSize(
226 colXx=self.colXx,
227 colYy=self.colYy,
228 colXy=self.colXy,
229 sizeType=self.sizeType,
230 ),
231 )
233 # distortion-type ellipticity has a shear response of 2, so we need to
234 # divide by 2 so that the rho-stats do not depend on the
235 # ellipticity-type.
236 # Note: For distortion, the responsitivity is 2(1 - e^2_{rms}),
237 # where e_rms is the root mean square ellipticity per component.
238 # This is expected to be small and we ignore it here.
239 # This definition of responsitivity is consistent with the definions
240 # used in the rho-statistics calculations for the HSC shear catalog
241 # papers (Mandelbaum et al. 2018, Li et al., 2022).
242 responsitivity = 2.0 if self.ellipticityType == "distortion" else 1.0
244 # Call the actions on the data.
245 eMEAS = calcEMeas(data, **kwargs)
246 if self.ellipticityType == "distortion":
247 _LOG.debug("Correction value of responsitivity would be %f", 2 - np.mean(np.abs(eMEAS) ** 2))
248 eMEAS /= responsitivity # type: ignore
249 e1, e2 = np.real(eMEAS), np.imag(eMEAS)
250 eRes = calcEDiff(data, **kwargs)
251 eRes /= responsitivity # type: ignore
252 e1Res, e2Res = np.real(eRes), np.imag(eRes)
253 sizeRes = -1 * calcSizeResidual(data, **kwargs) # sign flip residual to T_psf - T_model
255 # Scale the sizeRes by ellipticities
256 e1SizeRes = e1 * sizeRes
257 e2SizeRes = e2 * sizeRes
259 # Package the arguments to capture auto-/cross-correlations for the
260 # Rho statistics.
261 args = {
262 0: (sizeRes, None),
263 1: (e1Res, e2Res, None, None),
264 2: (e1, e2, e1Res, e2Res),
265 3: (e1SizeRes, e2SizeRes, None, None),
266 4: (e1Res, e2Res, e1SizeRes, e2SizeRes),
267 5: (e1, e2, e1SizeRes, e2SizeRes),
268 }
270 ra: Vector = data[self.colRa] # type: ignore
271 dec: Vector = data[self.colDec] # type: ignore
273 treecorr_config_dict = self.treecorr.toDict()
275 # Swap rng_seed with an rng instance in treecorr config.
276 rng = np.random.RandomState(treecorr_config_dict.pop("rng_seed"))
277 treecorr_config_dict["rng"] = rng
279 # Pass the appropriate arguments to the correlator and build a dict
280 rhoStats: Mapping[str, treecorr.BinnedCorr2] = {}
281 for rhoIndex in range(1, 6):
282 _LOG.info("Calculating rho-%d", rhoIndex)
283 rhoStats[f"rho{rhoIndex}"] = self._corrSpin2( # type: ignore[index]
284 ra,
285 dec,
286 *(args[rhoIndex]),
287 treecorr_config_dict=treecorr_config_dict,
288 )
290 _LOG.info("Calculating rho3alt")
291 rhoStats["rho3alt"] = self._corrSpin0( # type: ignore[index]
292 ra,
293 dec,
294 *(args[0]),
295 treecorr_config_dict=treecorr_config_dict,
296 )
297 return cast(KeyedData, rhoStats)
299 @classmethod
300 def _corrSpin0(
301 cls,
302 ra: Vector,
303 dec: Vector,
304 k1: Vector,
305 k2: Vector | None = None,
306 raUnits: str = "degrees",
307 decUnits: str = "degrees",
308 treecorr_config_dict: Mapping[str, Any] | None = None,
309 ) -> KKCorrelation:
310 """Function to compute correlations between at most two scalar fields.
312 This is used to compute rho3alt statistics, given the appropriate
313 spin-0 (scalar) fields, usually fractional size residuals.
315 Parameters
316 ----------
317 ra : `numpy.array`
318 The right ascension values of entries in the catalog.
319 dec : `numpy.array`
320 The declination values of entries in the catalog.
321 k1 : `numpy.array`
322 The primary scalar field.
323 k2 : `numpy.array`, optional
324 The secondary scalar field.
325 Autocorrelation of the primary field is computed if `None`.
326 raUnits : `str`, optional
327 Unit of the right ascension values. Valid options are
328 "degrees", "arcmin", "arcsec", "hours" or "radians".
329 decUnits : `str`, optional
330 Unit of the declination values. Valid options are
331 "degrees", "arcmin", "arcsec", "hours" or "radians".
332 treecorr_config_dict: `dict`, optional
333 Config dictionary to be passed to `treecorr`
334 (`treecorr.KKCorrelation` or `treecorr.Catalog`).
336 Returns
337 -------
338 xy : `treecorr.KKCorrelation`
339 A `treecorr.KKCorrelation` object containing the correlation
340 function.
341 """
342 _LOG.debug(
343 "No. of entries: %d. The number of pairs in the resulting KKCorrelation cannot exceed %d",
344 len(ra),
345 len(ra) * (len(ra) - 1) / 2,
346 )
347 xy = treecorr.KKCorrelation(config=treecorr_config_dict)
348 catA = treecorr.Catalog(
349 config=treecorr_config_dict,
350 ra=ra,
351 dec=dec,
352 k=k1,
353 ra_units=raUnits,
354 dec_units=decUnits,
355 logger=_LOG,
356 )
357 if k2 is None:
358 # Calculate the auto-correlation
359 xy.process(catA)
360 else:
361 catB = treecorr.Catalog(
362 config=treecorr_config_dict,
363 ra=ra,
364 dec=dec,
365 k=k2,
366 ra_units=raUnits,
367 dec_units=decUnits,
368 logger=_LOG,
369 patch_centers=catA.patch_centers,
370 )
371 # Calculate the cross-correlation
372 xy.process(catA, catB)
374 _LOG.debug("Correlated %d pairs based on the config set.", sum(xy.npairs))
375 return xy
377 @classmethod
378 def _corrSpin2(
379 cls,
380 ra: Vector,
381 dec: Vector,
382 g1a: Vector,
383 g2a: Vector,
384 g1b: Vector | None = None,
385 g2b: Vector | None = None,
386 raUnits: str = "degrees",
387 decUnits: str = "degrees",
388 treecorr_config_dict: Mapping[str, Any] | None = None,
389 ) -> GGCorrelation:
390 """Function to compute correlations between shear-like fields.
392 This is used to compute Rho statistics, given the appropriate spin-2
393 (shear-like) fields.
395 Parameters
396 ----------
397 ra : `numpy.array`
398 The right ascension values of entries in the catalog.
399 dec : `numpy.array`
400 The declination values of entries in the catalog.
401 g1a : `numpy.array`
402 The first component of the primary shear-like field.
403 g2a : `numpy.array`
404 The second component of the primary shear-like field.
405 g1b : `numpy.array`, optional
406 The first component of the secondary shear-like field.
407 Autocorrelation of the primary field is computed if `None`.
408 g2b : `numpy.array`, optional
409 The second component of the secondary shear-like field.
410 Autocorrelation of the primary field is computed if `None`.
411 raUnits : `str`, optional
412 Unit of the right ascension values. Valid options are
413 "degrees", "arcmin", "arcsec", "hours" or "radians".
414 decUnits : `str`, optional
415 Unit of the declination values. Valid options are
416 "degrees", "arcmin", "arcsec", "hours" or "radians".
417 treecorr_config_dict : `dict`, optional
418 Config dictionary to be passed to `treecorr`
419 (`treecorr.GGCorrelation` or `treecorr.Catalog`).
421 Returns
422 -------
423 xy : `treecorr.GGCorrelation`
424 A `treecorr.GGCorrelation` object containing the correlation
425 function.
426 """
427 _LOG.debug(
428 "No. of entries: %d. The number of pairs in the resulting GGCorrelation cannot exceed %d",
429 len(ra),
430 len(ra) * (len(ra) - 1) / 2,
431 )
432 xy = treecorr.GGCorrelation(config=treecorr_config_dict)
433 catA = treecorr.Catalog(
434 config=treecorr_config_dict,
435 ra=ra,
436 dec=dec,
437 g1=g1a,
438 g2=g2a,
439 ra_units=raUnits,
440 dec_units=decUnits,
441 logger=_LOG,
442 )
443 if g1b is None or g2b is None:
444 # Calculate the auto-correlation
445 xy.process(catA)
446 else:
447 catB = treecorr.Catalog(
448 config=treecorr_config_dict,
449 ra=ra,
450 dec=dec,
451 g1=g1b,
452 g2=g2b,
453 ra_units=raUnits,
454 dec_units=decUnits,
455 logger=_LOG,
456 patch_centers=catA.patch_centers,
457 )
458 # Calculate the cross-correlation
459 xy.process(catA, catB)
461 _LOG.debug("Correlated %d pairs based on the config set.", sum(xy.npairs))
462 return xy