forked from feast-dev/feast
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathfeature_store.py
More file actions
2119 lines (1841 loc) · 85 KB
/
Copy pathfeature_store.py
File metadata and controls
2119 lines (1841 loc) · 85 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
# Copyright 2019 The Feast Authors
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import copy
import itertools
import os
import warnings
from collections import Counter, defaultdict
from datetime import datetime
from pathlib import Path
from typing import (
TYPE_CHECKING,
Any,
Dict,
Iterable,
List,
Mapping,
Optional,
Sequence,
Set,
Tuple,
Union,
cast,
)
import pandas as pd
from colorama import Fore, Style
from google.protobuf.timestamp_pb2 import Timestamp
from tqdm import tqdm
from feast import feature_server, flags, flags_helper, utils
from feast.base_feature_view import BaseFeatureView
from feast.data_source import DataSource
from feast.diff.infra_diff import InfraDiff, diff_infra_protos
from feast.diff.registry_diff import RegistryDiff, apply_diff_to_registry, diff_between
from feast.entity import Entity
from feast.errors import (
EntityNotFoundException,
ExperimentalFeatureNotEnabled,
FeatureNameCollisionError,
FeatureViewNotFoundException,
RequestDataNotFoundInEntityDfException,
RequestDataNotFoundInEntityRowsException,
)
from feast.feast_object import FeastObject
from feast.feature_service import FeatureService
from feast.feature_view import (
DUMMY_ENTITY,
DUMMY_ENTITY_ID,
DUMMY_ENTITY_NAME,
DUMMY_ENTITY_VAL,
FeatureView,
)
from feast.inference import (
update_data_sources_with_inferred_event_timestamp_col,
update_entities_with_inferred_types_from_feature_views,
update_feature_views_with_inferred_features,
)
from feast.infra.infra_object import Infra
from feast.infra.provider import Provider, RetrievalJob, get_provider
from feast.on_demand_feature_view import OnDemandFeatureView
from feast.online_response import OnlineResponse
from feast.protos.feast.core.InfraObject_pb2 import Infra as InfraProto
from feast.protos.feast.serving.ServingService_pb2 import (
FieldStatus,
GetOnlineFeaturesResponse,
)
from feast.protos.feast.types.EntityKey_pb2 import EntityKey as EntityKeyProto
from feast.protos.feast.types.Value_pb2 import RepeatedValue, Value
from feast.registry import Registry
from feast.repo_config import RepoConfig, load_repo_config
from feast.repo_contents import RepoContents
from feast.request_feature_view import RequestFeatureView
from feast.saved_dataset import SavedDataset, SavedDatasetStorage
from feast.type_map import python_values_to_proto_values
from feast.usage import log_exceptions, log_exceptions_and_usage, set_usage_attribute
from feast.value_type import ValueType
from feast.version import get_version
warnings.simplefilter("once", DeprecationWarning)
if TYPE_CHECKING:
from feast.embedded_go.online_features_service import EmbeddedOnlineFeatureServer
class FeatureStore:
"""
A FeatureStore object is used to define, create, and retrieve features.
Args:
repo_path (optional): Path to a `feature_store.yaml` used to configure the
feature store.
config (optional): Configuration object used to configure the feature store.
"""
config: RepoConfig
repo_path: Path
_registry: Registry
_provider: Provider
_go_server: Optional["EmbeddedOnlineFeatureServer"]
@log_exceptions
def __init__(
self, repo_path: Optional[str] = None, config: Optional[RepoConfig] = None,
):
"""
Creates a FeatureStore object.
Raises:
ValueError: If both or neither of repo_path and config are specified.
"""
if repo_path is not None and config is not None:
raise ValueError("You cannot specify both repo_path and config.")
if config is not None:
self.repo_path = Path(os.getcwd())
self.config = config
elif repo_path is not None:
self.repo_path = Path(repo_path)
self.config = load_repo_config(Path(repo_path))
else:
raise ValueError("Please specify one of repo_path or config.")
registry_config = self.config.get_registry_config()
self._registry = Registry(registry_config, repo_path=self.repo_path)
self._registry._initialize_registry()
self._provider = get_provider(self.config, self.repo_path)
@log_exceptions
def version(self) -> str:
"""Returns the version of the current Feast SDK/CLI."""
return get_version()
@property
def registry(self) -> Registry:
"""Gets the registry of this feature store."""
return self._registry
@property
def project(self) -> str:
"""Gets the project of this feature store."""
return self.config.project
def _get_provider(self) -> Provider:
# TODO: Bake self.repo_path into self.config so that we dont only have one interface to paths
return self._provider
@log_exceptions_and_usage
def refresh_registry(self):
"""Fetches and caches a copy of the feature registry in memory.
Explicitly calling this method allows for direct control of the state of the registry cache. Every time this
method is called the complete registry state will be retrieved from the remote registry store backend
(e.g., GCS, S3), and the cache timer will be reset. If refresh_registry() is run before get_online_features()
is called, then get_online_features() will use the cached registry instead of retrieving (and caching) the
registry itself.
Additionally, the TTL for the registry cache can be set to infinity (by setting it to 0), which means that
refresh_registry() will become the only way to update the cached registry. If the TTL is set to a value
greater than 0, then once the cache becomes stale (more time than the TTL has passed), a new cache will be
downloaded synchronously, which may increase latencies if the triggering method is get_online_features().
"""
registry_config = self.config.get_registry_config()
registry = Registry(registry_config, repo_path=self.repo_path)
registry.refresh()
self._registry = registry
@log_exceptions_and_usage
def list_entities(self, allow_cache: bool = False) -> List[Entity]:
"""
Retrieves the list of entities from the registry.
Args:
allow_cache: Whether to allow returning entities from a cached registry.
Returns:
A list of entities.
"""
return self._list_entities(allow_cache)
def _list_entities(
self, allow_cache: bool = False, hide_dummy_entity: bool = True
) -> List[Entity]:
all_entities = self._registry.list_entities(
self.project, allow_cache=allow_cache
)
return [
entity
for entity in all_entities
if entity.name != DUMMY_ENTITY_NAME or not hide_dummy_entity
]
@log_exceptions_and_usage
def list_feature_services(self) -> List[FeatureService]:
"""
Retrieves the list of feature services from the registry.
Returns:
A list of feature services.
"""
return self._registry.list_feature_services(self.project)
@log_exceptions_and_usage
def list_feature_views(self, allow_cache: bool = False) -> List[FeatureView]:
"""
Retrieves the list of feature views from the registry.
Args:
allow_cache: Whether to allow returning entities from a cached registry.
Returns:
A list of feature views.
"""
return self._list_feature_views(allow_cache)
@log_exceptions_and_usage
def list_request_feature_views(
self, allow_cache: bool = False
) -> List[RequestFeatureView]:
"""
Retrieves the list of feature views from the registry.
Args:
allow_cache: Whether to allow returning entities from a cached registry.
Returns:
A list of feature views.
"""
return self._registry.list_request_feature_views(
self.project, allow_cache=allow_cache
)
def _list_feature_views(
self, allow_cache: bool = False, hide_dummy_entity: bool = True,
) -> List[FeatureView]:
feature_views = []
for fv in self._registry.list_feature_views(
self.project, allow_cache=allow_cache
):
if hide_dummy_entity and fv.entities[0] == DUMMY_ENTITY_NAME:
fv.entities = []
feature_views.append(fv)
return feature_views
@log_exceptions_and_usage
def list_on_demand_feature_views(
self, allow_cache: bool = False
) -> List[OnDemandFeatureView]:
"""
Retrieves the list of on demand feature views from the registry.
Returns:
A list of on demand feature views.
"""
return self._registry.list_on_demand_feature_views(
self.project, allow_cache=allow_cache
)
@log_exceptions_and_usage
def list_data_sources(self, allow_cache: bool = False) -> List[DataSource]:
"""
Retrieves the list of data sources from the registry.
Args:
allow_cache: Whether to allow returning data sources from a cached registry.
Returns:
A list of data sources.
"""
return self._registry.list_data_sources(self.project, allow_cache=allow_cache)
@log_exceptions_and_usage
def get_entity(self, name: str, allow_registry_cache: bool = False) -> Entity:
"""
Retrieves an entity.
Args:
name: Name of entity.
allow_registry_cache: (Optional) Whether to allow returning this entity from a cached registry
Returns:
The specified entity.
Raises:
EntityNotFoundException: The entity could not be found.
"""
return self._registry.get_entity(
name, self.project, allow_cache=allow_registry_cache
)
@log_exceptions_and_usage
def get_feature_service(
self, name: str, allow_cache: bool = False
) -> FeatureService:
"""
Retrieves a feature service.
Args:
name: Name of feature service.
allow_cache: Whether to allow returning feature services from a cached registry.
Returns:
The specified feature service.
Raises:
FeatureServiceNotFoundException: The feature service could not be found.
"""
return self._registry.get_feature_service(name, self.project, allow_cache)
@log_exceptions_and_usage
def get_feature_view(
self, name: str, allow_registry_cache: bool = False
) -> FeatureView:
"""
Retrieves a feature view.
Args:
name: Name of feature view.
allow_registry_cache: (Optional) Whether to allow returning this entity from a cached registry
Returns:
The specified feature view.
Raises:
FeatureViewNotFoundException: The feature view could not be found.
"""
return self._get_feature_view(name, allow_registry_cache=allow_registry_cache)
def _get_feature_view(
self,
name: str,
hide_dummy_entity: bool = True,
allow_registry_cache: bool = False,
) -> FeatureView:
feature_view = self._registry.get_feature_view(
name, self.project, allow_cache=allow_registry_cache
)
if hide_dummy_entity and feature_view.entities[0] == DUMMY_ENTITY_NAME:
feature_view.entities = []
return feature_view
@log_exceptions_and_usage
def get_on_demand_feature_view(self, name: str) -> OnDemandFeatureView:
"""
Retrieves a feature view.
Args:
name: Name of feature view.
Returns:
The specified feature view.
Raises:
FeatureViewNotFoundException: The feature view could not be found.
"""
return self._registry.get_on_demand_feature_view(name, self.project)
@log_exceptions_and_usage
def get_data_source(self, name: str) -> DataSource:
"""
Retrieves the list of data sources from the registry.
Args:
name: Name of the data source.
Returns:
The specified data source.
Raises:
DataSourceObjectNotFoundException: The data source could not be found.
"""
return self._registry.get_data_source(name, self.project)
@log_exceptions_and_usage
def delete_feature_view(self, name: str):
"""
Deletes a feature view.
Args:
name: Name of feature view.
Raises:
FeatureViewNotFoundException: The feature view could not be found.
"""
return self._registry.delete_feature_view(name, self.project)
@log_exceptions_and_usage
def delete_feature_service(self, name: str):
"""
Deletes a feature service.
Args:
name: Name of feature service.
Raises:
FeatureServiceNotFoundException: The feature view could not be found.
"""
return self._registry.delete_feature_service(name, self.project)
def _get_features(
self, features: Union[List[str], FeatureService], allow_cache: bool = False,
) -> List[str]:
_features = features
if not _features:
raise ValueError("No features specified for retrieval")
_feature_refs = []
if isinstance(_features, FeatureService):
feature_service_from_registry = self.get_feature_service(
_features.name, allow_cache
)
if feature_service_from_registry != _features:
warnings.warn(
"The FeatureService object that has been passed in as an argument is"
"inconsistent with the version from Registry. Potentially a newer version"
"of the FeatureService has been applied to the registry."
)
for projection in feature_service_from_registry.feature_view_projections:
_feature_refs.extend(
[
f"{projection.name_to_use()}:{f.name}"
for f in projection.features
]
)
else:
assert isinstance(_features, list)
_feature_refs = _features
return _feature_refs
def _should_use_plan(self):
"""Returns True if _plan and _apply_diffs should be used, False otherwise."""
# Currently only the local provider with sqlite online store supports _plan and _apply_diffs.
return self.config.provider == "local" and (
self.config.online_store and self.config.online_store.type == "sqlite"
)
def _validate_all_feature_views(
self,
views_to_update: List[FeatureView],
odfvs_to_update: List[OnDemandFeatureView],
request_views_to_update: List[RequestFeatureView],
):
"""Validates all feature views."""
if (
not flags_helper.enable_on_demand_feature_views(self.config)
and len(odfvs_to_update) > 0
):
raise ExperimentalFeatureNotEnabled(flags.FLAG_ON_DEMAND_TRANSFORM_NAME)
set_usage_attribute("odfv", bool(odfvs_to_update))
_validate_feature_views(
[*views_to_update, *odfvs_to_update, *request_views_to_update]
)
def _make_inferences(
self,
data_sources_to_update: List[DataSource],
entities_to_update: List[Entity],
views_to_update: List[FeatureView],
odfvs_to_update: List[OnDemandFeatureView],
):
"""Makes inferences for entities, feature views, and odfvs."""
update_entities_with_inferred_types_from_feature_views(
entities_to_update, views_to_update, self.config
)
update_data_sources_with_inferred_event_timestamp_col(
data_sources_to_update, self.config
)
update_data_sources_with_inferred_event_timestamp_col(
[view.batch_source for view in views_to_update], self.config
)
# New feature views may reference previously applied entities.
entities = self._list_entities()
update_feature_views_with_inferred_features(
views_to_update, entities + entities_to_update, self.config
)
for odfv in odfvs_to_update:
odfv.infer_features()
@log_exceptions_and_usage
def _plan(
self, desired_repo_contents: RepoContents
) -> Tuple[RegistryDiff, InfraDiff, Infra]:
"""Dry-run registering objects to metadata store.
The plan method dry-runs registering one or more definitions (e.g., Entity, FeatureView), and produces
a list of all the changes the that would be introduced in the feature repo. The changes computed by the plan
command are for informational purposes, and are not actually applied to the registry.
Args:
desired_repo_contents: The desired repo state.
Raises:
ValueError: The 'objects' parameter could not be parsed properly.
Examples:
Generate a plan adding an Entity and a FeatureView.
>>> from feast import FeatureStore, Entity, FeatureView, Feature, ValueType, FileSource, RepoConfig
>>> from feast.feature_store import RepoContents
>>> from datetime import timedelta
>>> fs = FeatureStore(repo_path="feature_repo")
>>> driver = Entity(name="driver_id", value_type=ValueType.INT64, description="driver id")
>>> driver_hourly_stats = FileSource(
... path="feature_repo/data/driver_stats.parquet",
... timestamp_field="event_timestamp",
... created_timestamp_column="created",
... )
>>> driver_hourly_stats_view = FeatureView(
... name="driver_hourly_stats",
... entities=["driver_id"],
... ttl=timedelta(seconds=86400 * 1),
... batch_source=driver_hourly_stats,
... )
>>> registry_diff, infra_diff, new_infra = fs._plan(RepoContents(
... data_sources={driver_hourly_stats},
... feature_views={driver_hourly_stats_view},
... on_demand_feature_views=set(),
... request_feature_views=set(),
... entities={driver},
... feature_services=set())) # register entity and feature view
"""
# Validate and run inference on all the objects to be registered.
self._validate_all_feature_views(
list(desired_repo_contents.feature_views),
list(desired_repo_contents.on_demand_feature_views),
list(desired_repo_contents.request_feature_views),
)
_validate_data_sources(list(desired_repo_contents.data_sources))
self._make_inferences(
list(desired_repo_contents.data_sources),
list(desired_repo_contents.entities),
list(desired_repo_contents.feature_views),
list(desired_repo_contents.on_demand_feature_views),
)
# Compute the desired difference between the current objects in the registry and
# the desired repo state.
registry_diff = diff_between(
self._registry, self.project, desired_repo_contents
)
# Compute the desired difference between the current infra, as stored in the registry,
# and the desired infra.
self._registry.refresh()
current_infra_proto = (
self._registry.cached_registry_proto.infra.__deepcopy__()
if self._registry.cached_registry_proto
else InfraProto()
)
desired_registry_proto = desired_repo_contents.to_registry_proto()
new_infra = self._provider.plan_infra(self.config, desired_registry_proto)
new_infra_proto = new_infra.to_proto()
infra_diff = diff_infra_protos(current_infra_proto, new_infra_proto)
return registry_diff, infra_diff, new_infra
@log_exceptions_and_usage
def _apply_diffs(
self, registry_diff: RegistryDiff, infra_diff: InfraDiff, new_infra: Infra
):
"""Applies the given diffs to the metadata store and infrastructure.
Args:
registry_diff: The diff between the current registry and the desired registry.
infra_diff: The diff between the current infra and the desired infra.
new_infra: The desired infra.
"""
infra_diff.update()
apply_diff_to_registry(
self._registry, registry_diff, self.project, commit=False
)
self._registry.update_infra(new_infra, self.project, commit=True)
@log_exceptions_and_usage
def apply(
self,
objects: Union[
DataSource,
Entity,
FeatureView,
OnDemandFeatureView,
RequestFeatureView,
FeatureService,
List[FeastObject],
],
objects_to_delete: Optional[List[FeastObject]] = None,
partial: bool = True,
):
"""Register objects to metadata store and update related infrastructure.
The apply method registers one or more definitions (e.g., Entity, FeatureView) and registers or updates these
objects in the Feast registry. Once the apply method has updated the infrastructure (e.g., create tables in
an online store), it will commit the updated registry. All operations are idempotent, meaning they can safely
be rerun.
Args:
objects: A single object, or a list of objects that should be registered with the Feature Store.
objects_to_delete: A list of objects to be deleted from the registry and removed from the
provider's infrastructure. This deletion will only be performed if partial is set to False.
partial: If True, apply will only handle the specified objects; if False, apply will also delete
all the objects in objects_to_delete, and tear down any associated cloud resources.
Raises:
ValueError: The 'objects' parameter could not be parsed properly.
Examples:
Register an Entity and a FeatureView.
>>> from feast import FeatureStore, Entity, FeatureView, Feature, ValueType, FileSource, RepoConfig
>>> from datetime import timedelta
>>> fs = FeatureStore(repo_path="feature_repo")
>>> driver = Entity(name="driver_id", value_type=ValueType.INT64, description="driver id")
>>> driver_hourly_stats = FileSource(
... path="feature_repo/data/driver_stats.parquet",
... timestamp_field="event_timestamp",
... created_timestamp_column="created",
... )
>>> driver_hourly_stats_view = FeatureView(
... name="driver_hourly_stats",
... entities=["driver_id"],
... ttl=timedelta(seconds=86400 * 1),
... batch_source=driver_hourly_stats,
... )
>>> fs.apply([driver_hourly_stats_view, driver]) # register entity and feature view
"""
# TODO: Add locking
if not isinstance(objects, Iterable):
objects = [objects]
assert isinstance(objects, list)
if not objects_to_delete:
objects_to_delete = []
# Separate all objects into entities, feature services, and different feature view types.
entities_to_update = [ob for ob in objects if isinstance(ob, Entity)]
views_to_update = [ob for ob in objects if isinstance(ob, FeatureView)]
request_views_to_update = [
ob for ob in objects if isinstance(ob, RequestFeatureView)
]
odfvs_to_update = [ob for ob in objects if isinstance(ob, OnDemandFeatureView)]
services_to_update = [ob for ob in objects if isinstance(ob, FeatureService)]
data_sources_set_to_update = {
ob for ob in objects if isinstance(ob, DataSource)
}
for fv in views_to_update:
data_sources_set_to_update.add(fv.batch_source)
if fv.stream_source:
data_sources_set_to_update.add(fv.stream_source)
if request_views_to_update:
warnings.warn(
"Request feature view is deprecated. "
"Please use request data source instead",
DeprecationWarning,
)
for rfv in request_views_to_update:
data_sources_set_to_update.add(rfv.request_data_source)
for odfv in odfvs_to_update:
for v in odfv.source_request_sources.values():
data_sources_set_to_update.add(v)
data_sources_to_update = list(data_sources_set_to_update)
# Validate all feature views and make inferences.
self._validate_all_feature_views(
views_to_update, odfvs_to_update, request_views_to_update
)
self._make_inferences(
data_sources_to_update, entities_to_update, views_to_update, odfvs_to_update
)
# Handle all entityless feature views by using DUMMY_ENTITY as a placeholder entity.
entities_to_update.append(DUMMY_ENTITY)
# Add all objects to the registry and update the provider's infrastructure.
for ds in data_sources_to_update:
self._registry.apply_data_source(ds, project=self.project, commit=False)
for view in itertools.chain(
views_to_update, odfvs_to_update, request_views_to_update
):
self._registry.apply_feature_view(view, project=self.project, commit=False)
for ent in entities_to_update:
self._registry.apply_entity(ent, project=self.project, commit=False)
for feature_service in services_to_update:
self._registry.apply_feature_service(
feature_service, project=self.project, commit=False
)
if not partial:
# Delete all registry objects that should not exist.
entities_to_delete = [
ob for ob in objects_to_delete if isinstance(ob, Entity)
]
views_to_delete = [
ob for ob in objects_to_delete if isinstance(ob, FeatureView)
]
request_views_to_delete = [
ob for ob in objects_to_delete if isinstance(ob, RequestFeatureView)
]
odfvs_to_delete = [
ob for ob in objects_to_delete if isinstance(ob, OnDemandFeatureView)
]
services_to_delete = [
ob for ob in objects_to_delete if isinstance(ob, FeatureService)
]
data_sources_to_delete = [
ob for ob in objects_to_delete if isinstance(ob, DataSource)
]
for data_source in data_sources_to_delete:
self._registry.delete_data_source(
data_source.name, project=self.project, commit=False
)
for entity in entities_to_delete:
self._registry.delete_entity(
entity.name, project=self.project, commit=False
)
for view in views_to_delete:
self._registry.delete_feature_view(
view.name, project=self.project, commit=False
)
for request_view in request_views_to_delete:
self._registry.delete_feature_view(
request_view.name, project=self.project, commit=False
)
for odfv in odfvs_to_delete:
self._registry.delete_feature_view(
odfv.name, project=self.project, commit=False
)
for service in services_to_delete:
self._registry.delete_feature_service(
service.name, project=self.project, commit=False
)
self._get_provider().update_infra(
project=self.project,
tables_to_delete=views_to_delete if not partial else [],
tables_to_keep=views_to_update,
entities_to_delete=entities_to_delete if not partial else [],
entities_to_keep=entities_to_update,
partial=partial,
)
self._registry.commit()
# go server needs to be reloaded to apply new configuration.
# we're stopping it here
# new server will be instantiated on the next online request
self._teardown_go_server()
@log_exceptions_and_usage
def teardown(self):
"""Tears down all local and cloud resources for the feature store."""
tables: List[FeatureView] = []
feature_views = self.list_feature_views()
tables.extend(feature_views)
entities = self.list_entities()
self._get_provider().teardown_infra(self.project, tables, entities)
self._registry.teardown()
self._teardown_go_server()
@log_exceptions_and_usage
def get_historical_features(
self,
entity_df: Union[pd.DataFrame, str],
features: Union[List[str], FeatureService],
full_feature_names: bool = False,
) -> RetrievalJob:
"""Enrich an entity dataframe with historical feature values for either training or batch scoring.
This method joins historical feature data from one or more feature views to an entity dataframe by using a time
travel join.
Each feature view is joined to the entity dataframe using all entities configured for the respective feature
view. All configured entities must be available in the entity dataframe. Therefore, the entity dataframe must
contain all entities found in all feature views, but the individual feature views can have different entities.
Time travel is based on the configured TTL for each feature view. A shorter TTL will limit the
amount of scanning that will be done in order to find feature data for a specific entity key. Setting a short
TTL may result in null values being returned.
Args:
entity_df (Union[pd.DataFrame, str]): An entity dataframe is a collection of rows containing all entity
columns (e.g., customer_id, driver_id) on which features need to be joined, as well as a event_timestamp
column used to ensure point-in-time correctness. Either a Pandas DataFrame can be provided or a string
SQL query. The query must be of a format supported by the configured offline store (e.g., BigQuery)
features: The list of features that should be retrieved from the offline store. These features can be
specified either as a list of string feature references or as a feature service. String feature
references must have format "feature_view:feature", e.g. "customer_fv:daily_transactions".
full_feature_names: If True, feature names will be prefixed with the corresponding feature view name,
changing them from the format "feature" to "feature_view__feature" (e.g. "daily_transactions"
changes to "customer_fv__daily_transactions").
Returns:
RetrievalJob which can be used to materialize the results.
Raises:
ValueError: Both or neither of features and feature_refs are specified.
Examples:
Retrieve historical features from a local offline store.
>>> from feast import FeatureStore, RepoConfig
>>> import pandas as pd
>>> fs = FeatureStore(repo_path="feature_repo")
>>> entity_df = pd.DataFrame.from_dict(
... {
... "driver_id": [1001, 1002],
... "event_timestamp": [
... datetime(2021, 4, 12, 10, 59, 42),
... datetime(2021, 4, 12, 8, 12, 10),
... ],
... }
... )
>>> retrieval_job = fs.get_historical_features(
... entity_df=entity_df,
... features=[
... "driver_hourly_stats:conv_rate",
... "driver_hourly_stats:acc_rate",
... "driver_hourly_stats:avg_daily_trips",
... ],
... )
>>> feature_data = retrieval_job.to_df()
"""
_feature_refs = self._get_features(features)
(
all_feature_views,
all_request_feature_views,
all_on_demand_feature_views,
) = self._get_feature_views_to_use(features)
if all_request_feature_views:
warnings.warn(
"Request feature view is deprecated. "
"Please use request data source instead",
DeprecationWarning,
)
# TODO(achal): _group_feature_refs returns the on demand feature views, but it's no passed into the provider.
# This is a weird interface quirk - we should revisit the `get_historical_features` to
# pass in the on demand feature views as well.
fvs, odfvs, request_fvs, request_fv_refs = _group_feature_refs(
_feature_refs,
all_feature_views,
all_request_feature_views,
all_on_demand_feature_views,
)
feature_views = list(view for view, _ in fvs)
on_demand_feature_views = list(view for view, _ in odfvs)
request_feature_views = list(view for view, _ in request_fvs)
set_usage_attribute("odfv", bool(on_demand_feature_views))
set_usage_attribute("request_fv", bool(request_feature_views))
# Check that the right request data is present in the entity_df
if type(entity_df) == pd.DataFrame:
entity_pd_df = cast(pd.DataFrame, entity_df)
for fv in request_feature_views:
for feature in fv.features:
if feature.name not in entity_pd_df.columns:
raise RequestDataNotFoundInEntityDfException(
feature_name=feature.name, feature_view_name=fv.name
)
for odfv in on_demand_feature_views:
odfv_request_data_schema = odfv.get_request_data_schema()
for feature_name in odfv_request_data_schema.keys():
if feature_name not in entity_pd_df.columns:
raise RequestDataNotFoundInEntityDfException(
feature_name=feature_name, feature_view_name=odfv.name,
)
_validate_feature_refs(_feature_refs, full_feature_names)
# Drop refs that refer to RequestFeatureViews since they don't need to be fetched and
# already exist in the entity_df
_feature_refs = [ref for ref in _feature_refs if ref not in request_fv_refs]
provider = self._get_provider()
job = provider.get_historical_features(
self.config,
feature_views,
_feature_refs,
entity_df,
self._registry,
self.project,
full_feature_names,
)
return job
@log_exceptions_and_usage
def create_saved_dataset(
self,
from_: RetrievalJob,
name: str,
storage: SavedDatasetStorage,
tags: Optional[Dict[str, str]] = None,
feature_service: Optional[FeatureService] = None,
) -> SavedDataset:
"""
Execute provided retrieval job and persist its outcome in given storage.
Storage type (eg, BigQuery or Redshift) must be the same as globally configured offline store.
After data successfully persisted saved dataset object with dataset metadata is committed to the registry.
Name for the saved dataset should be unique within project, since it's possible to overwrite previously stored dataset
with the same name.
Returns:
SavedDataset object with attached RetrievalJob
Raises:
ValueError if given retrieval job doesn't have metadata
"""
warnings.warn(
"Saving dataset is an experimental feature. "
"This API is unstable and it could and most probably will be changed in the future. "
"We do not guarantee that future changes will maintain backward compatibility.",
RuntimeWarning,
)
if not from_.metadata:
raise ValueError(
"RetrievalJob must contains metadata. "
"Use RetrievalJob produced by get_historical_features"
)
dataset = SavedDataset(
name=name,
features=from_.metadata.features,
join_keys=from_.metadata.keys,
full_feature_names=from_.full_feature_names,
storage=storage,
tags=tags,
feature_service_name=feature_service.name if feature_service else None,
)
dataset.min_event_timestamp = from_.metadata.min_event_timestamp
dataset.max_event_timestamp = from_.metadata.max_event_timestamp
from_.persist(storage)
dataset = dataset.with_retrieval_job(
self._get_provider().retrieve_saved_dataset(
config=self.config, dataset=dataset
)
)
self._registry.apply_saved_dataset(dataset, self.project, commit=True)
return dataset
@log_exceptions_and_usage
def get_saved_dataset(self, name: str) -> SavedDataset:
"""
Find a saved dataset in the registry by provided name and
create a retrieval job to pull whole dataset from storage (offline store).
If dataset couldn't be found by provided name SavedDatasetNotFound exception will be raised.
Data will be retrieved from globally configured offline store.
Returns:
SavedDataset with RetrievalJob attached
Raises:
SavedDatasetNotFound
"""
warnings.warn(
"Retrieving datasets is an experimental feature. "
"This API is unstable and it could and most probably will be changed in the future. "
"We do not guarantee that future changes will maintain backward compatibility.",
RuntimeWarning,
)
dataset = self._registry.get_saved_dataset(name, self.project)
provider = self._get_provider()
retrieval_job = provider.retrieve_saved_dataset(