Coverage for python/lsst/pipe/base/graph/_implDetails.py: 15%
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
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
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/>.
21from __future__ import annotations
23__all__ = ("_DatasetTracker", "DatasetTypeName", "_pruner")
25from collections import defaultdict
26from itertools import chain
27from typing import DefaultDict, Dict, Generic, Iterable, NewType, Optional, Set, TypeVar
29import networkx as nx
30from lsst.daf.butler import DatasetRef, NamedKeyDict, Quantum
31from lsst.pipe.base.connections import AdjustQuantumHelper
33from .._status import NoWorkFound
34from ..pipeline import TaskDef
35from .quantumNode import QuantumNode
37# NewTypes
38DatasetTypeName = NewType("DatasetTypeName", str)
40# Generic type parameters
41_T = TypeVar("_T", DatasetTypeName, DatasetRef)
42_U = TypeVar("_U", TaskDef, QuantumNode)
45class _DatasetTracker(Generic[_T, _U]):
46 r"""This is a generic container for tracking keys which are produced or
47 consumed by some value. In the context of a QuantumGraph, keys may be
48 `~lsst.daf.butler.DatasetRef`\ s and the values would be Quanta that either
49 produce or consume those `~lsst.daf.butler.DatasetRef`\ s.
51 Prameters
52 ---------
53 createInverse : bool
54 When adding a key associated with a producer or consumer, also create
55 and inverse mapping that allows looking up all the keys associated with
56 some value. Defaults to False.
57 """
59 def __init__(self, createInverse: bool = False):
60 self._producers: Dict[_T, _U] = {}
61 self._consumers: DefaultDict[_T, Set[_U]] = defaultdict(set)
62 self._createInverse = createInverse
63 if self._createInverse:
64 self._itemsDict: DefaultDict[_U, Set[_T]] = defaultdict(set)
66 def addProducer(self, key: _T, value: _U):
67 """Add a key which is produced by some value.
69 Parameters
70 ----------
71 key : TypeVar
72 The type to track
73 value : TypeVar
74 The type associated with the production of the key
76 Raises
77 ------
78 ValueError
79 Raised if key is already declared to be produced by another value
80 """
81 if (existing := self._producers.get(key)) is not None and existing != value:
82 raise ValueError(f"Only one node is allowed to produce {key}, the current producer is {existing}")
83 self._producers[key] = value
84 if self._createInverse:
85 self._itemsDict[value].add(key)
87 def removeProducer(self, key: _T, value: _U):
88 """Remove a value (e.g. QuantumNode or TaskDef) from being considered
89 a producer of the corresponding key. It is not an error to remove a
90 key that is not in the tracker.
92 Parameters
93 ----------
94 key : TypeVar
95 The type to track
96 value : TypeVar
97 The type associated with the production of the key
98 """
99 self._producers.pop(key, None)
100 if self._createInverse:
101 if result := self._itemsDict.get(value):
102 result.discard(key)
104 def addConsumer(self, key: _T, value: _U):
105 """Add a key which is consumed by some value.
107 Parameters
108 ----------
109 key : TypeVar
110 The type to track
111 value : TypeVar
112 The type associated with the consumption of the key
113 """
114 self._consumers[key].add(value)
115 if self._createInverse:
116 self._itemsDict[value].add(key)
118 def removeConsumer(self, key: _T, value: _U):
119 """Remove a value (e.g. QuantumNode or TaskDef) from being considered
120 a consumer of the corresponding key. It is not an error to remove a
121 key that is not in the tracker.
123 Parameters
124 ----------
125 key : TypeVar
126 The type to track
127 value : TypeVar
128 The type associated with the consumption of the key
129 """
130 result = self._consumers.get(key)
131 if result := self._consumers.get(key):
132 result.discard(value)
133 if self._createInverse:
134 if result := self._itemsDict.get(value):
135 result.discard(key)
137 def getConsumers(self, key: _T) -> set[_U]:
138 """Return all values associated with the consumption of the supplied
139 key.
141 Parameters
142 ----------
143 key : TypeVar
144 The type which has been tracked in the _DatasetTracker
145 """
146 return self._consumers.get(key, set())
148 def getProducer(self, key: _T) -> Optional[_U]:
149 """Return the value associated with the consumption of the supplied
150 key.
152 Parameters
153 ----------
154 key : TypeVar
155 The type which has been tracked in the _DatasetTracker
156 """
157 # This tracker may have had all nodes associated with a key removed
158 # and if there are no refs (empty set) should return None
159 return producer if (producer := self._producers.get(key)) else None
161 def getAll(self, key: _T) -> set[_U]:
162 """Return all consumers and the producer associated with the the
163 supplied key.
165 Parameters
166 ----------
167 key : TypeVar
168 The type which has been tracked in the _DatasetTracker
169 """
171 return self.getConsumers(key).union(x for x in (self.getProducer(key),) if x is not None)
173 @property
174 def inverse(self) -> Optional[DefaultDict[_U, Set[_T]]]:
175 """Return the inverse mapping if class was instantiated to create an
176 inverse, else return None.
177 """
178 return self._itemsDict if self._createInverse else None
180 def makeNetworkXGraph(self) -> nx.DiGraph:
181 """Create a NetworkX graph out of all the contained keys, using the
182 relations of producer and consumers to create the edges.
184 Returns:
185 graph : networkx.DiGraph
186 The graph created out of the supplied keys and their relations
187 """
188 graph = nx.DiGraph()
189 for entry in self._producers.keys() | self._consumers.keys():
190 producer = self.getProducer(entry)
191 consumers = self.getConsumers(entry)
192 # This block is for tasks that consume existing inputs
193 if producer is None and consumers:
194 for consumer in consumers:
195 graph.add_node(consumer)
196 # This block is for tasks that produce output that is not consumed
197 # in this graph
198 elif producer is not None and not consumers:
199 graph.add_node(producer)
200 # all other connections
201 else:
202 for consumer in consumers:
203 graph.add_edge(producer, consumer)
204 return graph
206 def keys(self) -> Set[_T]:
207 """Return all tracked keys."""
208 return self._producers.keys() | self._consumers.keys()
210 def remove(self, key: _T):
211 """Remove a key and its corresponding value from the tracker, this is
212 a no-op if the key is not in the tracker.
214 Parameters
215 ----------
216 key : TypeVar
217 A key tracked by the DatasetTracker
218 """
219 self._producers.pop(key, None)
220 self._consumers.pop(key, None)
222 def __contains__(self, key: _T) -> bool:
223 """Check if a key is in the _DatasetTracker
225 Parameters
226 ----------
227 key : TypeVar
228 The key to check
230 Returns
231 -------
232 contains : bool
233 Boolean of the presence of the supplied key
234 """
235 return key in self._producers or key in self._consumers
238def _pruner(
239 datasetRefDict: _DatasetTracker[DatasetRef, QuantumNode],
240 refsToRemove: Iterable[DatasetRef],
241 *,
242 alreadyPruned: Optional[Set[QuantumNode]] = None,
243):
244 r"""Prune supplied dataset refs out of datasetRefDict container, recursing
245 to additional nodes dependant on pruned refs. This function modifies
246 datasetRefDict in-place.
248 Parameters
249 ----------
250 datasetRefDict : `_DatasetTracker[DatasetRef, QuantumNode]`
251 The dataset tracker that maps `DatasetRef`\ s to the Quantum Nodes
252 that produce/consume that `DatasetRef`
253 refsToRemove : `Iterable` of `DatasetRef`
254 The `DatasetRef`\ s which should be pruned from the input dataset
255 tracker
256 alreadyPruned : `set` of `QuantumNode`
257 A set of nodes which have been pruned from the dataset tracker
258 """
259 if alreadyPruned is None:
260 alreadyPruned = set()
261 for ref in refsToRemove:
262 # make a copy here, because this structure will be modified in
263 # recursion, hitting a node more than once won't be much of an
264 # issue, as we skip anything that has been processed
265 nodes = set(datasetRefDict.getConsumers(ref))
266 for node in nodes:
267 # This node will never be associated with this ref
268 datasetRefDict.removeConsumer(ref, node)
269 if node in alreadyPruned:
270 continue
271 # find the connection corresponding to the input ref
272 connectionRefs = node.quantum.inputs.get(ref.datasetType)
273 if connectionRefs is None:
274 # look to see if any inputs are component refs that match the
275 # input ref to prune
276 others = ref.datasetType.makeAllComponentDatasetTypes()
277 # for each other component type check if there are assocated
278 # refs
279 for other in others:
280 connectionRefs = node.quantum.inputs.get(other)
281 if connectionRefs is not None:
282 # now search the component refs and see which one
283 # matches the ref to trim
284 for cr in connectionRefs:
285 if cr.makeCompositeRef() == ref:
286 toRemove = cr
287 break
288 else:
289 # Ref must be an initInput ref and we want to ignore those
290 raise RuntimeError(f"Cannot prune on non-Input dataset type {ref.datasetType.name}")
291 else:
292 toRemove = ref
294 tmpRefs = set(connectionRefs).difference((toRemove,))
295 tmpConnections = NamedKeyDict(node.quantum.inputs.items())
296 tmpConnections[toRemove.datasetType] = list(tmpRefs)
297 helper = AdjustQuantumHelper(inputs=tmpConnections, outputs=node.quantum.outputs)
298 assert node.quantum.dataId is not None, (
299 "assert to make the type checker happy, it should not "
300 "actually be possible to not have dataId set to None "
301 "at this point"
302 )
304 # Try to adjust the quantum with the reduced refs to make sure the
305 # node will still satisfy all its conditions.
306 #
307 # If it can't because NoWorkFound is raised, that means a
308 # connection is no longer present, and the node should be removed
309 # from the graph.
310 try:
311 helper.adjust_in_place(node.taskDef.connections, node.taskDef.label, node.quantum.dataId)
312 newQuantum = Quantum(
313 taskName=node.quantum.taskName,
314 taskClass=node.quantum.taskClass,
315 dataId=node.quantum.dataId,
316 initInputs=node.quantum.initInputs,
317 inputs=helper.inputs,
318 outputs=helper.outputs,
319 )
320 # If the inputs or outputs were adjusted to something different
321 # than what was supplied by the graph builder, dissassociate
322 # node from those refs, and if they are output refs, prune them
323 # from downstream tasks. This means that based on new inputs
324 # the task wants to produce fewer outputs, or consume fewer
325 # inputs.
326 for condition, existingMapping, newMapping, remover in (
327 (
328 helper.inputs_adjusted,
329 node.quantum.inputs,
330 helper.inputs,
331 datasetRefDict.removeConsumer,
332 ),
333 (
334 helper.outputs_adjusted,
335 node.quantum.outputs,
336 helper.outputs,
337 datasetRefDict.removeProducer,
338 ),
339 ):
340 if condition:
341 notNeeded = set()
342 for key in existingMapping:
343 if key not in newMapping:
344 compositeRefs = (
345 r if not r.isComponent() else r.makeCompositeRef()
346 for r in existingMapping[key]
347 )
348 notNeeded |= set(compositeRefs)
349 continue
350 notNeeded |= set(existingMapping[key]) - set(newMapping[key])
351 if notNeeded:
352 for ref in notNeeded:
353 if ref.isComponent():
354 ref = ref.makeCompositeRef()
355 remover(ref, node)
356 if remover is datasetRefDict.removeProducer:
357 _pruner(datasetRefDict, notNeeded, alreadyPruned=alreadyPruned)
358 object.__setattr__(node, "quantum", newQuantum)
359 noWorkFound = False
361 except NoWorkFound:
362 noWorkFound = True
364 if noWorkFound:
365 # This will throw if the length is less than the minimum number
366 for tmpRef in chain(
367 chain.from_iterable(node.quantum.inputs.values()), node.quantum.initInputs.values()
368 ):
369 if tmpRef.isComponent():
370 tmpRef = tmpRef.makeCompositeRef()
371 datasetRefDict.removeConsumer(tmpRef, node)
372 alreadyPruned.add(node)
373 # prune all outputs produced by this node
374 # mark that none of these will be produced
375 forwardPrunes = set()
376 for forwardRef in chain.from_iterable(node.quantum.outputs.values()):
377 datasetRefDict.removeProducer(forwardRef, node)
378 forwardPrunes.add(forwardRef)
379 _pruner(datasetRefDict, forwardPrunes, alreadyPruned=alreadyPruned)