Coverage for python/lsst/ap/association/mapApData.py : 23%

Hot-keys on this page
r m x p toggle line displays
j k next/prev highlighted chunk
0 (zero) top of page
1 (one) first highlighted chunk
1#
2# Developed for the LSST Data Management System.
3# This product includes software developed by the LSST Project
4# (http://www.lsst.org).
5# See the COPYRIGHT file at the top-level directory of this distribution
6# for details of code ownership.
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 GNU General Public License
19# along with this program. If not, see <http://www.gnu.org/licenses/>.
20#
22"""Classes for taking science pipeline outputs and creating data products for
23use in ap_association and the alert production database (APDB).
24"""
26__all__ = ["MapApDataConfig", "MapApDataTask",
27 "MapDiaSourceConfig", "MapDiaSourceTask",
28 "UnpackApdbFlags"]
30import numpy as np
31import os
32import yaml
34import lsst.afw.table as afwTable
35from lsst.daf.base import DateTime
36import lsst.pipe.base as pipeBase
37import lsst.pex.config as pexConfig
38from lsst.pex.exceptions import RuntimeError
39from lsst.utils import getPackageDir
40from .afwUtils import make_dia_source_schema
43class MapApDataConfig(pexConfig.Config):
44 """Configuration for the generic MapApDataTask class.
45 """
46 copyColumns = pexConfig.DictField(
47 keytype=str,
48 itemtype=str,
49 doc="Mapping of input SciencePipelines columns to output DPDD "
50 "columns.",
51 default={"id": "id",
52 "parent": "parent",
53 "coord_ra": "coord_ra",
54 "coord_dec": "coord_dec"}
55 )
58class MapApDataTask(pipeBase.Task):
59 """Generic mapper class for copying values from a science pipelines catalog
60 into a product for use in ap_association or the APDB.
61 """
62 ConfigClass = MapApDataConfig
63 _DefaultName = "mapApDataTask"
65 def __init__(self, inputSchema, outputSchema, **kwargs):
66 pipeBase.Task.__init__(self, **kwargs)
67 self.inputSchema = inputSchema
68 self.outputSchema = outputSchema
70 self.mapper = afwTable.SchemaMapper(inputSchema, outputSchema)
72 for inputName, outputName in self.config.copyColumns.items():
73 self.mapper.addMapping(
74 self.inputSchema.find(inputName).key,
75 outputName,
76 True)
78 @pipeBase.timeMethod
79 def run(self, inputCatalog, exposure=None):
80 """Copy data from the inputCatalog into an output catalog with
81 requested columns.
83 Parameters
84 ----------
85 inputCatalog: `lsst.afw.table.SourceCatalog`
86 Input catalog with data to be copied into new output catalog.
88 Returns
89 -------
90 outputCatalog: `lsst.afw.table.SourceCatalog`
91 Output catalog with data copied from input and new column names.
92 """
93 outputCatalog = afwTable.SourceCatalog(self.outputSchema)
94 outputCatalog.extend(inputCatalog, self.mapper)
96 if not outputCatalog.isContiguous():
97 raise RuntimeError("Output catalogs must be contiguous.")
99 return outputCatalog
102class MapDiaSourceConfig(pexConfig.Config):
103 """Config for the DiaSourceMapperTask
104 """
105 copyColumns = pexConfig.DictField(
106 keytype=str,
107 itemtype=str,
108 doc="Mapping of input SciencePipelines columns to output DPDD "
109 "columns.",
110 default={"id": "id",
111 "parent": "parent",
112 "coord_ra": "coord_ra",
113 "coord_dec": "coord_dec",
114 "slot_Centroid_x": "x",
115 "slot_Centroid_xErr": "xErr",
116 "slot_Centroid_y": "y",
117 "slot_Centroid_yErr": "yErr",
118 "slot_ApFlux_instFlux": "apFlux",
119 "slot_ApFlux_instFluxErr": "apFluxErr",
120 "slot_PsfFlux_instFlux": "psFlux",
121 "slot_PsfFlux_instFluxErr": "psFluxErr",
122 "ip_diffim_DipoleFit_orientation": "dipAngle",
123 "ip_diffim_DipoleFit_chi2dof": "dipChi2",
124 "ip_diffim_forced_PsfFlux_instFlux": "totFlux",
125 "ip_diffim_forced_PsfFlux_instFluxErr": "totFluxErr",
126 "ip_diffim_DipoleFit_flag_classification": "isDipole",
127 "slot_Shape_xx": "ixx",
128 "slot_Shape_xxErr": "ixxErr",
129 "slot_Shape_yy": "iyy",
130 "slot_Shape_yyErr": "iyyErr",
131 "slot_Shape_xy": "ixy",
132 "slot_Shape_xyErr": "ixyErr",
133 "slot_PsfShape_xx": "ixxPSF",
134 "slot_PsfShape_yy": "iyyPSF",
135 "slot_PsfShape_xy": "ixyPSF"}
136 )
137 calibrateColumns = pexConfig.ListField(
138 dtype=str,
139 doc="Flux columns in the input catalog to calibrate.",
140 default=["slot_ApFlux", "slot_PsfFlux", "ip_diffim_forced_PsfFlux"]
141 )
142 flagMap = pexConfig.Field(
143 dtype=str,
144 doc="Yaml file specifying SciencePipelines flag fields to bit packs.",
145 default=os.path.join(getPackageDir("ap_association"),
146 "data",
147 "association-flag-map.yaml"),
148 )
149 dipFluxPrefix = pexConfig.Field(
150 dtype=str,
151 doc="Prefix of the Dipole measurement column containing negative and "
152 "positive flux lobes.",
153 default="ip_diffim_DipoleFit",
154 )
155 dipSepColumn = pexConfig.Field(
156 dtype=str,
157 doc="Column of the separation of the negative and positive poles of "
158 "the dipole.",
159 default="ip_diffim_DipoleFit_separation"
160 )
163class MapDiaSourceTask(MapApDataTask):
164 """Task specific for copying columns from science pipelines catalogs,
165 calibrating them, for use in ap_association and the APDB.
167 This task also copies information from the exposure such as the ExpsoureId
168 and the exposure date as specified in the DPDD.
169 """
171 ConfigClass = MapDiaSourceConfig
172 _DefaultName = "mapDiaSourceTask"
174 def __init__(self, inputSchema, **kwargs):
175 MapApDataTask.__init__(self,
176 inputSchema=inputSchema,
177 outputSchema=make_dia_source_schema(),
178 **kwargs)
179 self._create_bit_pack_mappings()
181 def _create_bit_pack_mappings(self):
182 """Setup all flag bit packings.
183 """
184 self.bit_pack_columns = []
185 with open(self.config.flagMap) as yaml_stream:
186 table_list = list(yaml.safe_load_all(yaml_stream))
187 for table in table_list:
188 if table['tableName'] == 'DiaSource':
189 self.bit_pack_columns = table['columns']
190 break
192 # Test that all flags requested are present in both the input and
193 # output schemas.
194 for outputFlag in self.bit_pack_columns:
195 try:
196 self.outputSchema.find(outputFlag['columnName'])
197 except KeyError:
198 raise KeyError(
199 "Requested column %s not found in MapDiaSourceTask output "
200 "schema. Please check that the requested output column "
201 "exists." % outputFlag['columnName'])
202 bitList = outputFlag['bitList']
203 for bit in bitList:
204 try:
205 self.inputSchema.find(bit['name'])
206 except KeyError:
207 raise KeyError(
208 "Requested column %s not found in MapDiaSourceTask input "
209 "schema. Please check that the requested input column "
210 "exists." % outputFlag['columnName'])
212 @pipeBase.timeMethod
213 def run(self, inputCatalog, exposure, return_pandas=False):
214 """Copy data from the inputCatalog into an output catalog with
215 requested columns.
217 Parameters
218 ----------
219 inputCatalog : `lsst.afw.table.SourceCatalog`
220 Input catalog with data to be copied into new output catalog.
221 exposure: `lsst.afw.image.Exposure`
222 Exposure with containing the PhotoCalib object relevant to this
223 catalog.
224 return_pandas : `bool`
225 Return `pandas.DataFrame` instead of `lsst.afw.table.SourceCatalog`
227 Returns
228 -------
229 outputCatalog: `lsst.afw.table.SourceCatalog` or `pandas.DataFrame`
230 Output catalog with data copied from input and new column names.
231 """
232 visit_info = exposure.getInfo().getVisitInfo()
233 ccdVisitId = visit_info.getExposureId()
234 midPointTaiMJD = visit_info.getDate().get(system=DateTime.MJD)
235 filterId = exposure.getFilter().getId()
236 # TODO DM-27170: fix this [0] workaround which gets a single character
237 # representation of the band.
238 filterName = exposure.getFilter().getCanonicalName()[0]
239 wcs = exposure.getWcs()
241 photoCalib = exposure.getPhotoCalib()
243 outputCatalog = afwTable.SourceCatalog(self.outputSchema)
244 outputCatalog.reserve(len(inputCatalog))
246 for inputRecord in inputCatalog:
247 outputRecord = outputCatalog.addNew()
248 outputRecord.assign(inputRecord, self.mapper)
249 self.calibrateFluxes(inputRecord, outputRecord, photoCalib)
250 self.computeDipoleFluxes(inputRecord, outputRecord, photoCalib)
251 self.computeDipoleSep(inputRecord, outputRecord, wcs)
252 self.bitPackFlags(inputRecord, outputRecord)
253 self.computeBBoxSize(inputRecord, outputRecord)
254 outputRecord.set("ccdVisitId", ccdVisitId)
255 outputRecord.set("midPointTai", midPointTaiMJD)
256 outputRecord.set("filterId", filterId)
257 outputRecord.set("filterName", filterName)
259 if not outputCatalog.isContiguous():
260 raise RuntimeError("Output catalogs must be contiguous.")
262 if return_pandas:
263 return self._convert_to_pandas(outputCatalog)
264 return outputCatalog
266 def calibrateFluxes(self, inputRecord, outputRecord, photoCalib):
267 """Copy flux values into an output record and calibrate them.
269 Parameters
270 ----------
271 inputRecord : `lsst.afw.table.SourceRecord`
272 Record to copy flux values from.
273 outputRecord : `lsst.afw.table.SourceRecord`
274 Record to copy and calibrate values into.
275 photoCalib : `lsst.afw.image.PhotoCalib`
276 Calibration object from the difference exposure.
277 """
278 for col_name in self.config.calibrateColumns:
279 meas = photoCalib.instFluxToNanojansky(inputRecord, col_name)
280 outputRecord.set(self.config.copyColumns[col_name + "_instFlux"],
281 meas.value)
282 outputRecord.set(
283 self.config.copyColumns[col_name + "_instFluxErr"],
284 meas.error)
286 def computeDipoleFluxes(self, inputRecord, outputRecord, photoCalib):
287 """Calibrate and compute dipole mean flux and diff flux.
289 Parameters
290 ----------
291 inputRecord : `lsst.afw.table.SourceRecord`
292 Record to copy flux values from.
293 outputRecord : `lsst.afw.table.SourceRecord`
294 Record to copy and calibrate values into.
295 photoCalib `lsst.afw.image.PhotoCalib`
296 Calibration object from the difference exposure.
297 """
299 neg_meas = photoCalib.instFluxToNanojansky(
300 inputRecord, self.config.dipFluxPrefix + "_neg")
301 pos_meas = photoCalib.instFluxToNanojansky(
302 inputRecord, self.config.dipFluxPrefix + "_pos")
303 outputRecord.set(
304 "dipMeanFlux",
305 0.5 * (np.abs(neg_meas.value) + np.abs(pos_meas.value)))
306 outputRecord.set(
307 "dipMeanFluxErr",
308 0.5 * np.sqrt(neg_meas.error ** 2 + pos_meas.error ** 2))
309 outputRecord.set(
310 "dipFluxDiff",
311 np.abs(pos_meas.value) - np.abs(neg_meas.value))
312 outputRecord.set(
313 "dipFluxDiffErr",
314 np.sqrt(neg_meas.error ** 2 + pos_meas.error ** 2))
316 def computeDipoleSep(self, inputRecord, outputRecord, wcs):
317 """Convert the dipole separation from pixels to arcseconds.
319 Parameters
320 ----------
321 inputRecord : `lsst.afw.table.SourceRecord`
322 Record to copy flux values from.
323 outputRecord : `lsst.afw.table.SourceRecord`
324 Record to copy and calibrate values into.
325 wcs : `lsst.afw.geom.SkyWcs`
326 Wcs of image inputRecords was observed.
327 """
328 pixScale = wcs.getPixelScale(inputRecord.getCentroid())
329 dipSep = pixScale * inputRecord.get(self.config.dipSepColumn)
330 outputRecord.set("dipLength", dipSep.asArcseconds())
332 def bitPackFlags(self, inputRecord, outputRecord):
333 """Pack requested flag columns in inputRecord into single columns in
334 outputRecord.
336 Parameters
337 ----------
338 inputRecord : `lsst.afw.table.SourceRecord`
339 Record to copy flux values from.
340 outputRecord : `lsst.afw.table.SourceRecord`
341 Record to copy and calibrate values into.
342 """
343 for outputFlag in self.bit_pack_columns:
344 bitList = outputFlag['bitList']
345 value = 0
346 for bit in bitList:
347 value += inputRecord[bit['name']] * 2 ** bit['bit']
348 outputRecord.set(outputFlag['columnName'], value)
350 def computeBBoxSize(self, inputRecord, outputRecord):
351 """Compute the size of a square bbox that fully contains the detection
352 footprint.
354 Parameters
355 ----------
356 inputRecord : `lsst.afw.table.SourceRecord`
357 Record to copy flux values from.
358 outputRecord : `lsst.afw.table.SourceRecord`
359 Record to copy and calibrate values into.
360 """
361 footprintBBox = inputRecord.getFootprint().getBBox()
362 # Compute twice the size of the largest dimension of the footprint
363 # bounding box. This is the largest footprint we should need to cover
364 # the complete DiaSource assuming the centroid is withing the bounding
365 # box.
366 maxSize = 2 * np.max([footprintBBox.getWidth(),
367 footprintBBox.getHeight()])
368 recX = inputRecord.getCentroid().x
369 recY = inputRecord.getCentroid().y
370 bboxSize = int(
371 np.ceil(2 * np.max(np.fabs([footprintBBox.maxX - recX,
372 footprintBBox.minX - recX,
373 footprintBBox.maxY - recY,
374 footprintBBox.minY - recY]))))
375 if bboxSize > maxSize:
376 bboxSize = maxSize
377 outputRecord.set("bboxSize", bboxSize)
379 def _convert_to_pandas(self, inputCatalog):
380 """Convert input afw table to pandas.
382 Using afwTable.toAstropy().to_pandas() alone is not sufficient to
383 properly store data in the Apdb. We must also convert the RA/DEC values
384 from radians to degrees and rename several columns.
386 Parameters
387 ----------
388 inputCatalog : `lsst.afw.table.SourceCatalog`
389 Catalog to convert to panads and rename columns.
391 Returns
392 -------
393 catalog : `pandas.DataFrame`
394 """
395 catalog = inputCatalog.asAstropy().to_pandas()
396 catalog.rename(columns={"coord_ra": "ra",
397 "coord_dec": "decl",
398 "id": "diaSourceId",
399 "parent": "parentDiaSourceId"},
400 inplace=True)
401 catalog["ra"] = np.degrees(catalog["ra"])
402 catalog["decl"] = np.degrees(catalog["decl"])
404 return catalog
407class UnpackApdbFlags:
408 """Class for unpacking bits from integer flag fields stored in the Apdb.
410 Attributes
411 ----------
412 flag_map_file : `str`
413 Absolute or relative path to a yaml file specifiying mappings of flags
414 to integer bits.
415 table_name : `str`
416 Name of the Apdb table the integer bit data are coming from.
417 """
419 def __init__(self, flag_map_file, table_name):
420 self.bit_pack_columns = []
421 with open(flag_map_file) as yaml_stream:
422 table_list = list(yaml.safe_load_all(yaml_stream))
423 for table in table_list:
424 if table['tableName'] == table_name:
425 self.bit_pack_columns = table['columns']
426 break
428 self.output_flag_columns = {}
430 for column in self.bit_pack_columns:
431 names = []
432 for bit in column["bitList"]:
433 names.append((bit["name"], np.bool))
434 self.output_flag_columns[column["columnName"]] = names
436 def unpack(self, input_flag_values, flag_name):
437 """Determine individual boolean flags from an input array of unsigned
438 ints.
440 Parameters
441 ----------
442 input_flag_values : array-like of type uint
443 Input integer flags to unpack.
444 flag_name : `str`
445 Apdb column name of integer flags to unpack. Names of packed int
446 flags are given by the flag_map_file.
448 Returns
449 -------
450 output_flags : `numpy.ndarray`
451 Numpy named tuple of booleans.
452 """
453 bit_names_types = self.output_flag_columns[flag_name]
454 output_flags = np.zeros(len(input_flag_values), dtype=bit_names_types)
456 for bit_idx, (bit_name, dtypes) in enumerate(bit_names_types):
457 masked_bits = np.bitwise_and(input_flag_values, 2 ** bit_idx)
458 output_flags[bit_name] = masked_bits
460 return output_flags