LSSTApplications  18.1.0
LSSTDataManagementBasePackage
reserveSourcesTask.py
Go to the documentation of this file.
1 #
2 # LSST Data Management System
3 #
4 # Copyright 2008-2017 AURA/LSST.
5 #
6 # This product includes software developed by the
7 # LSST Project (http://www.lsst.org/).
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 LSST License Statement and
20 # the GNU General Public License along with this program. If not,
21 # see <https://www.lsstcorp.org/LegalNotices/>.
22 #
23 
24 __all__ = ["ReserveSourcesConfig", "ReserveSourcesTask"]
25 
26 import numpy as np
27 
28 from lsst.pex.config import Config, Field
29 from lsst.pipe.base import Task, Struct
30 
31 
33  """Configuration for reserving sources"""
34  fraction = Field(dtype=float, default=0.0,
35  doc="Fraction of candidates to reserve from fitting; none if <= 0")
36  seed = Field(dtype=int, default=1,
37  doc=("This number will be added to the exposure ID to set the random seed for "
38  "reserving candidates"))
39 
40 
42  """Reserve sources from analysis
43 
44  We randomly select a fraction of sources that will be reserved
45  from analysis. This allows evaluation of the quality of model fits
46  using sources that were not involved in the fitting process.
47 
48  Constructor parameters
49  ----------------------
50  columnName : `str`, required
51  Name of flag column to add; we will suffix this with "_reserved".
52  schema : `lsst.afw.table.Schema`, required
53  Catalog schema.
54  doc : `str`
55  Documentation for column to add.
56  config : `ReserveSourcesConfig`
57  Configuration.
58  """
59  ConfigClass = ReserveSourcesConfig
60  _DefaultName = "reserveSources"
61 
62  def __init__(self, columnName=None, schema=None, doc=None, **kwargs):
63  Task.__init__(self, **kwargs)
64  assert columnName is not None, "columnName not provided"
65  assert schema is not None, "schema not provided"
66  self.columnName = columnName
67  self.key = schema.addField(self.columnName + "_reserved", type="Flag", doc=doc)
68 
69  def run(self, sources, prior=None, expId=0):
70  """Select sources to be reserved
71 
72  Reserved sources will be flagged in the catalog, and we will return
73  boolean arrays that identify the sources to be reserved from and
74  used in the analysis. Typically you'll want to use the sources
75  from the `use` array in your fitting, and use the sources from the
76  `reserved` array as an independent test of your fitting.
77 
78  Parameters
79  ----------
80  sources : `lsst.afw.table.Catalog` or `list` of `lsst.afw.table.Record`
81  Sources from which to select some to be reserved.
82  prior : `numpy.ndarray` of type `bool`, optional
83  Prior selection of sources. Should have the same length as
84  `sources`. If set, we will only consider for reservation sources
85  that are flagged `True` in this array.
86  expId : `int`
87  Exposure identifier; used for seeding the random number generator.
88 
89  Return struct contents
90  ----------------------
91  reserved : `numpy.ndarray` of type `bool`
92  Sources to be reserved are flagged `True` in this array.
93  use : `numpy.ndarray` of type `bool`
94  Sources the user should use in analysis are flagged `True`.
95  """
96  if prior is not None:
97  assert len(prior) == len(sources), "Length mismatch: %s vs %s" % (len(prior), len(sources))
98  numSources = prior.sum()
99  else:
100  numSources = len(sources)
101  selection = self.select(numSources, expId)
102  if prior is not None:
103  selection = self.applySelectionPrior(prior, selection)
104  self.markSources(sources, selection)
105  self.log.info("Reserved %d/%d sources", selection.sum(), len(selection))
106  return Struct(reserved=selection,
107  use=prior & ~selection if prior is not None else np.logical_not(selection))
108 
109  def select(self, numSources, expId=0):
110  """Randomly select some sources
111 
112  We return a boolean array with a random selection. The fraction
113  of sources selected is specified by the config parameter `fraction`.
114 
115  Parameters
116  ----------
117  numSources : `int`
118  Number of sources in catalog from which to select.
119  expId : `int`
120  Exposure identifier; used for seeding the random number generator.
121 
122  Returns
123  -------
124  selection : `numpy.ndarray` of type `bool`
125  Selected sources are flagged `True` in this array.
126  """
127  selection = np.zeros(numSources, dtype=bool)
128  if self.config.fraction <= 0:
129  return selection
130  reserve = int(np.round(numSources*self.config.fraction))
131  selection[:reserve] = True
132  rng = np.random.RandomState((self.config.seed + expId) & 0xFFFFFFFF)
133  rng.shuffle(selection)
134  return selection
135 
136  def applySelectionPrior(self, prior, selection):
137  """Apply selection to full catalog
138 
139  The `select` method makes a random selection of sources. If those
140  sources don't represent the full population (because a sub-selection
141  has already been made), then we need to generate a selection covering
142  the entire population.
143 
144  Parameters
145  ----------
146  prior : `numpy.ndarray` of type `bool`
147  Prior selection of sources, identifying the subset from which the
148  random selection has been made.
149  selection : `numpy.ndarray` of type `bool`
150  Selection of sources in subset identified by `prior`.
151 
152  Returns
153  -------
154  full : `numpy.ndarray` of type `bool`
155  Selection applied to full population.
156  """
157  full = np.zeros(len(prior), dtype=bool)
158  full[prior] = selection
159  return full
160 
161  def markSources(self, sources, selection):
162  """Mark sources in a list or catalog
163 
164  This requires iterating through the list and setting the flag in
165  each source individually. Even if the `sources` is a `Catalog`
166  with contiguous records, it's not currently possible to set a boolean
167  column (DM-6981) so we need to iterate.
168 
169  Parameters
170  ----------
171  catalog : `lsst.afw.table.Catalog` or `list` of `lsst.afw.table.Record`
172  Catalog in which to flag selected sources.
173  selection : `numpy.ndarray` of type `bool`
174  Selection of sources to mark.
175  """
176  for src, select in zip(sources, selection):
177  if select:
178  src.set(self.key, True)
def __init__(self, columnName=None, schema=None, doc=None, kwargs)