Coverage for python/lsst/pipe/base/pipelineTask.py: 68%

Shortcuts 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

17 statements  

1# This file is part of pipe_base. 

2# 

3# Developed for the LSST Data Management System. 

4# This product includes software developed by the LSST Project 

5# (http://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 <http://www.gnu.org/licenses/>. 

21 

22"""This module defines PipelineTask class and related methods. 

23""" 

24 

25__all__ = ["PipelineTask"] # Classes in this module 

26 

27from .butlerQuantumContext import ButlerQuantumContext 

28from .connections import InputQuantizedConnection, OutputQuantizedConnection 

29from .task import Task 

30 

31 

32class PipelineTask(Task): 

33 """Base class for all pipeline tasks. 

34 

35 This is an abstract base class for PipelineTasks which represents an 

36 algorithm executed by framework(s) on data which comes from data butler, 

37 resulting data is also stored in a data butler. 

38 

39 PipelineTask inherits from a `pipe.base.Task` and uses the same 

40 configuration mechanism based on `pex.config`. `PipelineTask` classes also 

41 have a `PipelineTaskConnections` class associated with their config which 

42 defines all of the IO a `PipelineTask` will need to do. PipelineTask 

43 sub-class typically implements `run()` method which receives Python-domain 

44 data objects and returns `pipe.base.Struct` object with resulting data. 

45 `run()` method is not supposed to perform any I/O, it operates entirely on 

46 in-memory objects. `runQuantum()` is the method (can be re-implemented in 

47 sub-class) where all necessary I/O is performed, it reads all input data 

48 from data butler into memory, calls `run()` method with that data, examines 

49 returned `Struct` object and saves some or all of that data back to data 

50 butler. `runQuantum()` method receives a `ButlerQuantumContext` instance to 

51 facilitate I/O, a `InputQuantizedConnection` instance which defines all 

52 input `lsst.daf.butler.DatasetRef`, and a `OutputQuantizedConnection` 

53 instance which defines all the output `lsst.daf.butler.DatasetRef` for a 

54 single invocation of PipelineTask. 

55 

56 Subclasses must be constructable with exactly the arguments taken by the 

57 PipelineTask base class constructor, but may support other signatures as 

58 well. 

59 

60 Attributes 

61 ---------- 

62 canMultiprocess : bool, True by default (class attribute) 

63 This class attribute is checked by execution framework, sub-classes 

64 can set it to ``False`` in case task does not support multiprocessing. 

65 

66 Parameters 

67 ---------- 

68 config : `pex.config.Config`, optional 

69 Configuration for this task (an instance of ``self.ConfigClass``, 

70 which is a task-specific subclass of `PipelineTaskConfig`). 

71 If not specified then it defaults to `self.ConfigClass()`. 

72 log : `logging.Logger`, optional 

73 Logger instance whose name is used as a log name prefix, or ``None`` 

74 for no prefix. 

75 initInputs : `dict`, optional 

76 A dictionary of objects needed to construct this PipelineTask, with 

77 keys matching the keys of the dictionary returned by 

78 `getInitInputDatasetTypes` and values equivalent to what would be 

79 obtained by calling `Butler.get` with those DatasetTypes and no data 

80 IDs. While it is optional for the base class, subclasses are 

81 permitted to require this argument. 

82 """ 

83 

84 canMultiprocess = True 

85 

86 def __init__(self, *, config=None, log=None, initInputs=None, **kwargs): 

87 super().__init__(config=config, log=log, **kwargs) 

88 

89 def run(self, **kwargs): 

90 """Run task algorithm on in-memory data. 

91 

92 This method should be implemented in a subclass. This method will 

93 receive keyword arguments whose names will be the same as names of 

94 connection fields describing input dataset types. Argument values will 

95 be data objects retrieved from data butler. If a dataset type is 

96 configured with ``multiple`` field set to ``True`` then the argument 

97 value will be a list of objects, otherwise it will be a single object. 

98 

99 If the task needs to know its input or output DataIds then it has to 

100 override `runQuantum` method instead. 

101 

102 This method should return a `Struct` whose attributes share the same 

103 name as the connection fields describing output dataset types. 

104 

105 Returns 

106 ------- 

107 struct : `Struct` 

108 Struct with attribute names corresponding to output connection 

109 fields 

110 

111 Examples 

112 -------- 

113 Typical implementation of this method may look like: 

114 

115 .. code-block:: python 

116 

117 def run(self, input, calib): 

118 # "input", "calib", and "output" are the names of the config 

119 # fields 

120 

121 # Assuming that input/calib datasets are `scalar` they are 

122 # simple objects, do something with inputs and calibs, produce 

123 # output image. 

124 image = self.makeImage(input, calib) 

125 

126 # If output dataset is `scalar` then return object, not list 

127 return Struct(output=image) 

128 

129 """ 

130 raise NotImplementedError("run() is not implemented") 

131 

132 def runQuantum( 

133 self, 

134 butlerQC: ButlerQuantumContext, 

135 inputRefs: InputQuantizedConnection, 

136 outputRefs: OutputQuantizedConnection, 

137 ): 

138 """Method to do butler IO and or transforms to provide in memory 

139 objects for tasks run method 

140 

141 Parameters 

142 ---------- 

143 butlerQC : `ButlerQuantumContext` 

144 A butler which is specialized to operate in the context of a 

145 `lsst.daf.butler.Quantum`. 

146 inputRefs : `InputQuantizedConnection` 

147 Datastructure whose attribute names are the names that identify 

148 connections defined in corresponding `PipelineTaskConnections` 

149 class. The values of these attributes are the 

150 `lsst.daf.butler.DatasetRef` objects associated with the defined 

151 input/prerequisite connections. 

152 outputRefs : `OutputQuantizedConnection` 

153 Datastructure whose attribute names are the names that identify 

154 connections defined in corresponding `PipelineTaskConnections` 

155 class. The values of these attributes are the 

156 `lsst.daf.butler.DatasetRef` objects associated with the defined 

157 output connections. 

158 """ 

159 inputs = butlerQC.get(inputRefs) 

160 outputs = self.run(**inputs) 

161 butlerQC.put(outputs, outputRefs) 

162 

163 def getResourceConfig(self): 

164 """Return resource configuration for this task. 

165 

166 Returns 

167 ------- 

168 Object of type `~config.ResourceConfig` or ``None`` if resource 

169 configuration is not defined for this task. 

170 """ 

171 return getattr(self.config, "resources", None)