Coverage for python/lsst/pipe/base/connections.py : 46%

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# 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/>.
22"""Module defining connection classes for PipelineTask.
23"""
25__all__ = ["PipelineTaskConnections", "InputQuantizedConnection", "OutputQuantizedConnection",
26 "DeferredDatasetRef", "iterConnections"]
28from collections import UserDict, namedtuple
29from types import SimpleNamespace
30import typing
32import itertools
33import string
35from . import config as configMod
36from .connectionTypes import (InitInput, InitOutput, Input, PrerequisiteInput,
37 Output, BaseConnection)
38from lsst.daf.butler import DatasetRef, DatasetType, NamedKeyDict, Quantum
40if typing.TYPE_CHECKING: 40 ↛ 41line 40 didn't jump to line 41, because the condition on line 40 was never true
41 from .config import PipelineTaskConfig
44class ScalarError(TypeError):
45 """Exception raised when dataset type is configured as scalar
46 but there are multiple data IDs in a Quantum for that dataset.
47 """
50class PipelineTaskConnectionDict(UserDict):
51 """This is a special dict class used by PipelineTaskConnectionMetaclass
53 This dict is used in PipelineTaskConnection class creation, as the
54 dictionary that is initially used as __dict__. It exists to
55 intercept connection fields declared in a PipelineTaskConnection, and
56 what name is used to identify them. The names are then added to class
57 level list according to the connection type of the class attribute. The
58 names are also used as keys in a class level dictionary associated with
59 the corresponding class attribute. This information is a duplicate of
60 what exists in __dict__, but provides a simple place to lookup and
61 iterate on only these variables.
62 """
63 def __init__(self, *args, **kwargs):
64 super().__init__(*args, **kwargs)
65 # Initialize class level variables used to track any declared
66 # class level variables that are instances of
67 # connectionTypes.BaseConnection
68 self.data['inputs'] = []
69 self.data['prerequisiteInputs'] = []
70 self.data['outputs'] = []
71 self.data['initInputs'] = []
72 self.data['initOutputs'] = []
73 self.data['allConnections'] = {}
75 def __setitem__(self, name, value):
76 if isinstance(value, Input):
77 self.data['inputs'].append(name)
78 elif isinstance(value, PrerequisiteInput):
79 self.data['prerequisiteInputs'].append(name)
80 elif isinstance(value, Output):
81 self.data['outputs'].append(name)
82 elif isinstance(value, InitInput): 82 ↛ 83line 82 didn't jump to line 83, because the condition on line 82 was never true
83 self.data['initInputs'].append(name)
84 elif isinstance(value, InitOutput): 84 ↛ 85line 84 didn't jump to line 85, because the condition on line 84 was never true
85 self.data['initOutputs'].append(name)
86 # This should not be an elif, as it needs tested for
87 # everything that inherits from BaseConnection
88 if isinstance(value, BaseConnection):
89 object.__setattr__(value, 'varName', name)
90 self.data['allConnections'][name] = value
91 # defer to the default behavior
92 super().__setitem__(name, value)
95class PipelineTaskConnectionsMetaclass(type):
96 """Metaclass used in the declaration of PipelineTaskConnections classes
97 """
98 def __prepare__(name, bases, **kwargs): # noqa: 805
99 # Create an instance of our special dict to catch and track all
100 # variables that are instances of connectionTypes.BaseConnection
101 # Copy any existing connections from a parent class
102 dct = PipelineTaskConnectionDict()
103 for base in bases:
104 if isinstance(base, PipelineTaskConnectionsMetaclass): 104 ↛ 103line 104 didn't jump to line 103, because the condition on line 104 was never false
105 for name, value in base.allConnections.items(): 105 ↛ 106line 105 didn't jump to line 106, because the loop on line 105 never started
106 dct[name] = value
107 return dct
109 def __new__(cls, name, bases, dct, **kwargs):
110 dimensionsValueError = TypeError("PipelineTaskConnections class must be created with a dimensions "
111 "attribute which is an iterable of dimension names")
113 if name != 'PipelineTaskConnections':
114 # Verify that dimensions are passed as a keyword in class
115 # declaration
116 if 'dimensions' not in kwargs: 116 ↛ 117line 116 didn't jump to line 117, because the condition on line 116 was never true
117 for base in bases:
118 if hasattr(base, 'dimensions'):
119 kwargs['dimensions'] = base.dimensions
120 break
121 if 'dimensions' not in kwargs:
122 raise dimensionsValueError
123 try:
124 dct['dimensions'] = set(kwargs['dimensions'])
125 except TypeError as exc:
126 raise dimensionsValueError from exc
127 # Lookup any python string templates that may have been used in the
128 # declaration of the name field of a class connection attribute
129 allTemplates = set()
130 stringFormatter = string.Formatter()
131 # Loop over all connections
132 for obj in dct['allConnections'].values():
133 nameValue = obj.name
134 # add all the parameters to the set of templates
135 for param in stringFormatter.parse(nameValue):
136 if param[1] is not None: 136 ↛ 137line 136 didn't jump to line 137, because the condition on line 136 was never true
137 allTemplates.add(param[1])
139 # look up any template from base classes and merge them all
140 # together
141 mergeDict = {}
142 for base in bases[::-1]:
143 if hasattr(base, 'defaultTemplates'): 143 ↛ 144line 143 didn't jump to line 144, because the condition on line 143 was never true
144 mergeDict.update(base.defaultTemplates)
145 if 'defaultTemplates' in kwargs: 145 ↛ 146line 145 didn't jump to line 146, because the condition on line 145 was never true
146 mergeDict.update(kwargs['defaultTemplates'])
148 if len(mergeDict) > 0: 148 ↛ 149line 148 didn't jump to line 149, because the condition on line 148 was never true
149 kwargs['defaultTemplates'] = mergeDict
151 # Verify that if templated strings were used, defaults were
152 # supplied as an argument in the declaration of the connection
153 # class
154 if len(allTemplates) > 0 and 'defaultTemplates' not in kwargs: 154 ↛ 155line 154 didn't jump to line 155, because the condition on line 154 was never true
155 raise TypeError("PipelineTaskConnection class contains templated attribute names, but no "
156 "defaut templates were provided, add a dictionary attribute named "
157 "defaultTemplates which contains the mapping between template key and value")
158 if len(allTemplates) > 0: 158 ↛ 160line 158 didn't jump to line 160, because the condition on line 158 was never true
159 # Verify all templates have a default, and throw if they do not
160 defaultTemplateKeys = set(kwargs['defaultTemplates'].keys())
161 templateDifference = allTemplates.difference(defaultTemplateKeys)
162 if templateDifference:
163 raise TypeError(f"Default template keys were not provided for {templateDifference}")
164 # Verify that templates do not share names with variable names
165 # used for a connection, this is needed because of how
166 # templates are specified in an associated config class.
167 nameTemplateIntersection = allTemplates.intersection(set(dct['allConnections'].keys()))
168 if len(nameTemplateIntersection) > 0:
169 raise TypeError(f"Template parameters cannot share names with Class attributes"
170 f" (conflicts are {nameTemplateIntersection}).")
171 dct['defaultTemplates'] = kwargs.get('defaultTemplates', {})
173 # Convert all the connection containers into frozensets so they cannot
174 # be modified at the class scope
175 for connectionName in ("inputs", "prerequisiteInputs", "outputs", "initInputs", "initOutputs"):
176 dct[connectionName] = frozenset(dct[connectionName])
177 # our custom dict type must be turned into an actual dict to be used in
178 # type.__new__
179 return super().__new__(cls, name, bases, dict(dct))
181 def __init__(cls, name, bases, dct, **kwargs):
182 # This overrides the default init to drop the kwargs argument. Python
183 # metaclasses will have this argument set if any kwargs are passes at
184 # class construction time, but should be consumed before calling
185 # __init__ on the type metaclass. This is in accordance with python
186 # documentation on metaclasses
187 super().__init__(name, bases, dct)
190class QuantizedConnection(SimpleNamespace):
191 """A Namespace to map defined variable names of connections to their
192 `lsst.daf.buter.DatasetRef`s
194 This class maps the names used to define a connection on a
195 PipelineTaskConnectionsClass to the corresponding
196 `lsst.daf.butler.DatasetRef`s provided by a `lsst.daf.butler.Quantum`
197 instance. This will be a quantum of execution based on the graph created
198 by examining all the connections defined on the
199 `PipelineTaskConnectionsClass`.
200 """
201 def __init__(self, **kwargs):
202 # Create a variable to track what attributes are added. This is used
203 # later when iterating over this QuantizedConnection instance
204 object.__setattr__(self, "_attributes", set())
206 def __setattr__(self, name: str, value: typing.Union[DatasetRef, typing.List[DatasetRef]]):
207 # Capture the attribute name as it is added to this object
208 self._attributes.add(name)
209 super().__setattr__(name, value)
211 def __delattr__(self, name):
212 object.__delattr__(self, name)
213 self._attributes.remove(name)
215 def __iter__(self) -> typing.Generator[typing.Tuple[str, typing.Union[DatasetRef,
216 typing.List[DatasetRef]]], None, None]:
217 """Make an Iterator for this QuantizedConnection
219 Iterating over a QuantizedConnection will yield a tuple with the name
220 of an attribute and the value associated with that name. This is
221 similar to dict.items() but is on the namespace attributes rather than
222 dict keys.
223 """
224 yield from ((name, getattr(self, name)) for name in self._attributes)
226 def keys(self) -> typing.Generator[str, None, None]:
227 """Returns an iterator over all the attributes added to a
228 QuantizedConnection class
229 """
230 yield from self._attributes
233class InputQuantizedConnection(QuantizedConnection):
234 pass
237class OutputQuantizedConnection(QuantizedConnection):
238 pass
241class DeferredDatasetRef(namedtuple("DeferredDatasetRefBase", "datasetRef")):
242 """Class which denotes that a datasetRef should be treated as deferred when
243 interacting with the butler
245 Parameters
246 ----------
247 datasetRef : `lsst.daf.butler.DatasetRef`
248 The `lsst.daf.butler.DatasetRef` that will be eventually used to
249 resolve a dataset
250 """
251 __slots__ = ()
254class PipelineTaskConnections(metaclass=PipelineTaskConnectionsMetaclass):
255 """PipelineTaskConnections is a class used to declare desired IO when a
256 PipelineTask is run by an activator
258 Parameters
259 ----------
260 config : `PipelineTaskConfig`
261 A `PipelineTaskConfig` class instance whose class has been configured
262 to use this `PipelineTaskConnectionsClass`
264 Notes
265 -----
266 ``PipelineTaskConnection`` classes are created by declaring class
267 attributes of types defined in `lsst.pipe.base.connectionTypes` and are
268 listed as follows:
270 * ``InitInput`` - Defines connections in a quantum graph which are used as
271 inputs to the ``__init__`` function of the `PipelineTask` corresponding
272 to this class
273 * ``InitOuput`` - Defines connections in a quantum graph which are to be
274 persisted using a butler at the end of the ``__init__`` function of the
275 `PipelineTask` corresponding to this class. The variable name used to
276 define this connection should be the same as an attribute name on the
277 `PipelineTask` instance. E.g. if an ``InitOutput`` is declared with
278 the name ``outputSchema`` in a ``PipelineTaskConnections`` class, then
279 a `PipelineTask` instance should have an attribute
280 ``self.outputSchema`` defined. Its value is what will be saved by the
281 activator framework.
282 * ``PrerequisiteInput`` - An input connection type that defines a
283 `lsst.daf.butler.DatasetType` that must be present at execution time,
284 but that will not be used during the course of creating the quantum
285 graph to be executed. These most often are things produced outside the
286 processing pipeline, such as reference catalogs.
287 * ``Input`` - Input `lsst.daf.butler.DatasetType` objects that will be used
288 in the ``run`` method of a `PipelineTask`. The name used to declare
289 class attribute must match a function argument name in the ``run``
290 method of a `PipelineTask`. E.g. If the ``PipelineTaskConnections``
291 defines an ``Input`` with the name ``calexp``, then the corresponding
292 signature should be ``PipelineTask.run(calexp, ...)``
293 * ``Output`` - A `lsst.daf.butler.DatasetType` that will be produced by an
294 execution of a `PipelineTask`. The name used to declare the connection
295 must correspond to an attribute of a `Struct` that is returned by a
296 `PipelineTask` ``run`` method. E.g. if an output connection is
297 defined with the name ``measCat``, then the corresponding
298 ``PipelineTask.run`` method must return ``Struct(measCat=X,..)`` where
299 X matches the ``storageClass`` type defined on the output connection.
301 The process of declaring a ``PipelineTaskConnection`` class involves
302 parameters passed in the declaration statement.
304 The first parameter is ``dimensions`` which is an iterable of strings which
305 defines the unit of processing the run method of a corresponding
306 `PipelineTask` will operate on. These dimensions must match dimensions that
307 exist in the butler registry which will be used in executing the
308 corresponding `PipelineTask`.
310 The second parameter is labeled ``defaultTemplates`` and is conditionally
311 optional. The name attributes of connections can be specified as python
312 format strings, with named format arguments. If any of the name parameters
313 on connections defined in a `PipelineTaskConnections` class contain a
314 template, then a default template value must be specified in the
315 ``defaultTemplates`` argument. This is done by passing a dictionary with
316 keys corresponding to a template identifier, and values corresponding to
317 the value to use as a default when formatting the string. For example if
318 ``ConnectionClass.calexp.name = '{input}Coadd_calexp'`` then
319 ``defaultTemplates`` = {'input': 'deep'}.
321 Once a `PipelineTaskConnections` class is created, it is used in the
322 creation of a `PipelineTaskConfig`. This is further documented in the
323 documentation of `PipelineTaskConfig`. For the purposes of this
324 documentation, the relevant information is that the config class allows
325 configuration of connection names by users when running a pipeline.
327 Instances of a `PipelineTaskConnections` class are used by the pipeline
328 task execution framework to introspect what a corresponding `PipelineTask`
329 will require, and what it will produce.
331 Examples
332 --------
333 >>> from lsst.pipe.base import connectionTypes as cT
334 >>> from lsst.pipe.base import PipelineTaskConnections
335 >>> from lsst.pipe.base import PipelineTaskConfig
336 >>> class ExampleConnections(PipelineTaskConnections,
337 ... dimensions=("A", "B"),
338 ... defaultTemplates={"foo": "Example"}):
339 ... inputConnection = cT.Input(doc="Example input",
340 ... dimensions=("A", "B"),
341 ... storageClass=Exposure,
342 ... name="{foo}Dataset")
343 ... outputConnection = cT.Output(doc="Example output",
344 ... dimensions=("A", "B"),
345 ... storageClass=Exposure,
346 ... name="{foo}output")
347 >>> class ExampleConfig(PipelineTaskConfig,
348 ... pipelineConnections=ExampleConnections):
349 ... pass
350 >>> config = ExampleConfig()
351 >>> config.connections.foo = Modified
352 >>> config.connections.outputConnection = "TotallyDifferent"
353 >>> connections = ExampleConnections(config=config)
354 >>> assert(connections.inputConnection.name == "ModifiedDataset")
355 >>> assert(connections.outputConnection.name == "TotallyDifferent")
356 """
358 def __init__(self, *, config: 'PipelineTaskConfig' = None):
359 self.inputs = set(self.inputs)
360 self.prerequisiteInputs = set(self.prerequisiteInputs)
361 self.outputs = set(self.outputs)
362 self.initInputs = set(self.initInputs)
363 self.initOutputs = set(self.initOutputs)
365 if config is None or not isinstance(config, configMod.PipelineTaskConfig):
366 raise ValueError("PipelineTaskConnections must be instantiated with"
367 " a PipelineTaskConfig instance")
368 self.config = config
369 # Extract the template names that were defined in the config instance
370 # by looping over the keys of the defaultTemplates dict specified at
371 # class declaration time
372 templateValues = {name: getattr(config.connections, name) for name in getattr(self,
373 'defaultTemplates').keys()}
374 # Extract the configured value corresponding to each connection
375 # variable. I.e. for each connection identifier, populate a override
376 # for the connection.name attribute
377 self._nameOverrides = {name: getattr(config.connections, name).format(**templateValues)
378 for name in self.allConnections.keys()}
380 # connections.name corresponds to a dataset type name, create a reverse
381 # mapping that goes from dataset type name to attribute identifier name
382 # (variable name) on the connection class
383 self._typeNameToVarName = {v: k for k, v in self._nameOverrides.items()}
385 def buildDatasetRefs(self, quantum: Quantum) -> typing.Tuple[InputQuantizedConnection,
386 OutputQuantizedConnection]:
387 """Builds QuantizedConnections corresponding to input Quantum
389 Parameters
390 ----------
391 quantum : `lsst.daf.butler.Quantum`
392 Quantum object which defines the inputs and outputs for a given
393 unit of processing
395 Returns
396 -------
397 retVal : `tuple` of (`InputQuantizedConnection`,
398 `OutputQuantizedConnection`) Namespaces mapping attribute names
399 (identifiers of connections) to butler references defined in the
400 input `lsst.daf.butler.Quantum`
401 """
402 inputDatasetRefs = InputQuantizedConnection()
403 outputDatasetRefs = OutputQuantizedConnection()
404 # operate on a reference object and an interable of names of class
405 # connection attributes
406 for refs, names in zip((inputDatasetRefs, outputDatasetRefs),
407 (itertools.chain(self.inputs, self.prerequisiteInputs), self.outputs)):
408 # get a name of a class connection attribute
409 for attributeName in names:
410 # get the attribute identified by name
411 attribute = getattr(self, attributeName)
412 # Branch if the attribute dataset type is an input
413 if attribute.name in quantum.predictedInputs:
414 # Get the DatasetRefs
415 quantumInputRefs = quantum.predictedInputs[attribute.name]
416 # if the dataset is marked to load deferred, wrap it in a
417 # DeferredDatasetRef
418 if attribute.deferLoad:
419 quantumInputRefs = [DeferredDatasetRef(datasetRef=ref) for ref in quantumInputRefs]
420 # Unpack arguments that are not marked multiples (list of
421 # length one)
422 if not attribute.multiple:
423 if len(quantumInputRefs) > 1:
424 raise ScalarError(
425 f"Received multiple datasets "
426 f"{', '.join(str(r.dataId) for r in quantumInputRefs)} "
427 f"for scalar connection {attributeName} "
428 f"({quantumInputRefs[0].datasetType.name}) "
429 f"of quantum for {quantum.taskName} with data ID {quantum.dataId}."
430 )
431 if len(quantumInputRefs) == 0:
432 continue
433 quantumInputRefs = quantumInputRefs[0]
434 # Add to the QuantizedConnection identifier
435 setattr(refs, attributeName, quantumInputRefs)
436 # Branch if the attribute dataset type is an output
437 elif attribute.name in quantum.outputs:
438 value = quantum.outputs[attribute.name]
439 # Unpack arguments that are not marked multiples (list of
440 # length one)
441 if not attribute.multiple:
442 value = value[0]
443 # Add to the QuantizedConnection identifier
444 setattr(refs, attributeName, value)
445 # Specified attribute is not in inputs or outputs dont know how
446 # to handle, throw
447 else:
448 raise ValueError(f"Attribute with name {attributeName} has no counterpoint "
449 "in input quantum")
450 return inputDatasetRefs, outputDatasetRefs
452 def adjustQuantum(self, datasetRefMap: NamedKeyDict[DatasetType, typing.Set[DatasetRef]]
453 ) -> NamedKeyDict[DatasetType, typing.Set[DatasetRef]]:
454 """Override to make adjustments to `lsst.daf.butler.DatasetRef` objects
455 in the `lsst.daf.butler.core.Quantum` during the graph generation stage
456 of the activator.
458 The base class implementation simply checks that input connections with
459 ``multiple`` set to `False` have no more than one dataset.
461 Parameters
462 ----------
463 datasetRefMap : `NamedKeyDict`
464 Mapping from dataset type to a `set` of
465 `lsst.daf.butler.DatasetRef` objects
467 Returns
468 -------
469 datasetRefMap : `NamedKeyDict`
470 Modified mapping of input with possibly adjusted
471 `lsst.daf.butler.DatasetRef` objects.
473 Raises
474 ------
475 ScalarError
476 Raised if any `Input` or `PrerequisiteInput` connection has
477 ``multiple`` set to `False`, but multiple datasets.
478 Exception
479 Overrides of this function have the option of raising an Exception
480 if a field in the input does not satisfy a need for a corresponding
481 pipelineTask, i.e. no reference catalogs are found.
482 """
483 for connection in itertools.chain(iterConnections(self, "inputs"),
484 iterConnections(self, "prerequisiteInputs")):
485 refs = datasetRefMap[connection.name]
486 if not connection.multiple and len(refs) > 1:
487 raise ScalarError(
488 f"Found multiple datasets {', '.join(str(r.dataId) for r in refs)} "
489 f"for scalar connection {connection.name} ({refs[0].datasetType.name})."
490 )
491 return datasetRefMap
494def iterConnections(connections: PipelineTaskConnections, connectionType: str) -> typing.Generator:
495 """Creates an iterator over the selected connections type which yields
496 all the defined connections of that type.
498 Parameters
499 ----------
500 connections: `PipelineTaskConnections`
501 An instance of a `PipelineTaskConnections` object that will be iterated
502 over.
503 connectionType: `str`
504 The type of connections to iterate over, valid values are inputs,
505 outputs, prerequisiteInputs, initInputs, initOutputs.
507 Yields
508 -------
509 connection: `BaseConnection`
510 A connection defined on the input connections object of the type
511 supplied. The yielded value Will be an derived type of
512 `BaseConnection`.
513 """
514 for name in getattr(connections, connectionType):
515 yield getattr(connections, name)