forked from openml/openml-python
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathfunctions.py
More file actions
609 lines (516 loc) 路 17.7 KB
/
Copy pathfunctions.py
File metadata and controls
609 lines (516 loc) 路 17.7 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
# License: BSD 3-Clause
# ruff: noqa: PLR0913
from __future__ import annotations
import warnings
from functools import partial
from typing import TYPE_CHECKING, Any
import pandas as pd
import xmltodict
import openml._api_calls
import openml.config
import openml.utils
from openml.study.study import OpenMLBenchmarkSuite, OpenMLStudy
if TYPE_CHECKING:
from openml.study.study import BaseStudy
def get_suite(suite_id: int | str) -> OpenMLBenchmarkSuite:
"""
Retrieves all relevant information of an OpenML benchmarking suite from the server.
Parameters
----------
study id : int, str
study id (numeric or alias)
Returns
-------
OpenMLSuite
The OpenML suite object
"""
study = _get_study(suite_id, entity_type="task")
assert isinstance(study, OpenMLBenchmarkSuite)
return study
def get_study(
study_id: int | str,
arg_for_backwards_compat: str | None = None, # noqa: ARG001
) -> OpenMLStudy: # F401
"""
Retrieves all relevant information of an OpenML study from the server.
Parameters
----------
study id : int, str
study id (numeric or alias)
arg_for_backwards_compat : str, optional
The example given in https://arxiv.org/pdf/1708.03731.pdf uses an older version of the
API which required specifying the type of study, i.e. tasks. We changed the
implementation of studies since then and split them up into suites (collections of tasks)
and studies (collections of runs) so this argument is no longer needed.
Returns
-------
OpenMLStudy
The OpenML study object
"""
if study_id == "OpenML100":
message = (
"It looks like you are running code from the OpenML100 paper. It still works, but lots "
"of things have changed since then. Please use `get_suite('OpenML100')` instead."
)
warnings.warn(message, DeprecationWarning, stacklevel=2)
openml.config.logger.warning(message)
study = _get_study(study_id, entity_type="task")
assert isinstance(study, OpenMLBenchmarkSuite)
return study # type: ignore
study = _get_study(study_id, entity_type="run")
assert isinstance(study, OpenMLStudy)
return study
def _get_study(id_: int | str, entity_type: str) -> BaseStudy:
xml_string = openml._api_calls._perform_api_call(f"study/{id_}", "get")
force_list_tags = (
"oml:data_id",
"oml:flow_id",
"oml:task_id",
"oml:setup_id",
"oml:run_id",
"oml:tag", # legacy.
)
result_dict = xmltodict.parse(xml_string, force_list=force_list_tags)["oml:study"]
study_id = int(result_dict["oml:id"])
alias = result_dict.get("oml:alias", None)
main_entity_type = result_dict["oml:main_entity_type"]
if entity_type != main_entity_type:
raise ValueError(
f"Unexpected entity type '{main_entity_type}' reported by the server"
f", expected '{entity_type}'"
)
benchmark_suite = result_dict.get("oml:benchmark_suite", None)
name = result_dict["oml:name"]
description = result_dict["oml:description"]
status = result_dict["oml:status"]
creation_date = result_dict["oml:creation_date"]
creator = result_dict["oml:creator"]
# tags is legacy. remove once no longer needed.
tags = []
if "oml:tag" in result_dict:
for tag in result_dict["oml:tag"]:
current_tag = {"name": tag["oml:name"], "write_access": tag["oml:write_access"]}
if "oml:window_start" in tag:
current_tag["window_start"] = tag["oml:window_start"]
tags.append(current_tag)
def get_nested_ids_from_result_dict(key: str, subkey: str) -> list[int] | None:
"""Extracts a list of nested IDs from a result dictionary.
Parameters
----------
key : str
Nested OpenML IDs.
subkey : str
The subkey contains the nested OpenML IDs.
Returns
-------
Optional[List]
A list of nested OpenML IDs, or None if the key is not present in the dictionary.
"""
if result_dict.get(key) is not None:
return [int(oml_id) for oml_id in result_dict[key][subkey]]
return None
datasets = get_nested_ids_from_result_dict("oml:data", "oml:data_id")
tasks = get_nested_ids_from_result_dict("oml:tasks", "oml:task_id")
if main_entity_type in ["runs", "run"]:
flows = get_nested_ids_from_result_dict("oml:flows", "oml:flow_id")
setups = get_nested_ids_from_result_dict("oml:setups", "oml:setup_id")
runs = get_nested_ids_from_result_dict("oml:runs", "oml:run_id")
study = OpenMLStudy(
study_id=study_id,
alias=alias,
benchmark_suite=benchmark_suite,
name=name,
description=description,
status=status,
creation_date=creation_date,
creator=creator,
tags=tags,
data=datasets,
tasks=tasks,
flows=flows,
setups=setups,
runs=runs,
) # type: BaseStudy
elif main_entity_type in ["tasks", "task"]:
study = OpenMLBenchmarkSuite(
suite_id=study_id,
alias=alias,
name=name,
description=description,
status=status,
creation_date=creation_date,
creator=creator,
tags=tags,
data=datasets,
tasks=tasks,
)
else:
raise ValueError(f"Unknown entity type {main_entity_type}")
return study
def create_study(
name: str,
description: str,
run_ids: list[int] | None = None,
alias: str | None = None,
benchmark_suite: int | None = None,
) -> OpenMLStudy:
"""
Creates an OpenML study (collection of data, tasks, flows, setups and run),
where the runs are the main entity (collection consists of runs and all
entities (flows, tasks, etc) that are related to these runs)
Parameters
----------
benchmark_suite : int (optional)
the benchmark suite (another study) upon which this study is ran.
name : str
the name of the study (meta-info)
description : str
brief description (meta-info)
run_ids : list, optional
a list of run ids associated with this study,
these can also be added later with ``attach_to_study``.
alias : str (optional)
a string ID, unique on server (url-friendly)
benchmark_suite: int (optional)
the ID of the suite for which this study contains run results
Returns
-------
OpenMLStudy
A local OpenML study object (call publish method to upload to server)
"""
return OpenMLStudy(
study_id=None,
alias=alias,
benchmark_suite=benchmark_suite,
name=name,
description=description,
status=None,
creation_date=None,
creator=None,
tags=None,
data=None,
tasks=None,
flows=None,
runs=run_ids if run_ids != [] else None,
setups=None,
)
def create_benchmark_suite(
name: str,
description: str,
task_ids: list[int],
alias: str | None = None,
) -> OpenMLBenchmarkSuite:
"""
Creates an OpenML benchmark suite (collection of entity types, where
the tasks are the linked entity)
Parameters
----------
name : str
the name of the study (meta-info)
description : str
brief description (meta-info)
task_ids : list
a list of task ids associated with this study
more can be added later with ``attach_to_suite``.
alias : str (optional)
a string ID, unique on server (url-friendly)
Returns
-------
OpenMLStudy
A local OpenML study object (call publish method to upload to server)
"""
return OpenMLBenchmarkSuite(
suite_id=None,
alias=alias,
name=name,
description=description,
status=None,
creation_date=None,
creator=None,
tags=None,
data=None,
tasks=task_ids,
)
def update_suite_status(suite_id: int, status: str) -> None:
"""
Updates the status of a study to either 'active' or 'deactivated'.
Parameters
----------
suite_id : int
The data id of the dataset
status : str,
'active' or 'deactivated'
"""
return update_study_status(suite_id, status)
def update_study_status(study_id: int, status: str) -> None:
"""
Updates the status of a study to either 'active' or 'deactivated'.
Parameters
----------
study_id : int
The data id of the dataset
status : str,
'active' or 'deactivated'
"""
legal_status = {"active", "deactivated"}
if status not in legal_status:
raise ValueError(f"Illegal status value. Legal values: {legal_status}")
data = {"study_id": study_id, "status": status} # type: openml._api_calls.DATA_TYPE
result_xml = openml._api_calls._perform_api_call("study/status/update", "post", data=data)
result = xmltodict.parse(result_xml)
server_study_id = result["oml:study_status_update"]["oml:id"]
server_status = result["oml:study_status_update"]["oml:status"]
if status != server_status or int(study_id) != int(server_study_id):
# This should never happen
raise ValueError("Study id/status does not collide")
def delete_suite(suite_id: int) -> bool:
"""Deletes a study from the OpenML server.
Parameters
----------
suite_id : int
OpenML id of the study
Returns
-------
bool
True iff the deletion was successful. False otherwise
"""
return delete_study(suite_id)
def delete_study(study_id: int) -> bool:
"""Deletes a study from the OpenML server.
Parameters
----------
study_id : int
OpenML id of the study
Returns
-------
bool
True iff the deletion was successful. False otherwise
"""
return openml.utils._delete_entity("study", study_id)
def attach_to_suite(suite_id: int, task_ids: list[int]) -> int:
"""Attaches a set of tasks to a benchmarking suite.
Parameters
----------
suite_id : int
OpenML id of the study
task_ids : list (int)
List of entities to link to the collection
Returns
-------
int
new size of the suite (in terms of explicitly linked entities)
"""
return attach_to_study(suite_id, task_ids)
def attach_to_study(study_id: int, run_ids: list[int]) -> int:
"""Attaches a set of runs to a study.
Parameters
----------
study_id : int
OpenML id of the study
run_ids : list (int)
List of entities to link to the collection
Returns
-------
int
new size of the study (in terms of explicitly linked entities)
"""
# Interestingly, there's no need to tell the server about the entity type, it knows by itself
result_xml = openml._api_calls._perform_api_call(
call=f"study/{study_id}/attach",
request_method="post",
data={"ids": ",".join(str(x) for x in run_ids)},
)
result = xmltodict.parse(result_xml)["oml:study_attach"]
return int(result["oml:linked_entities"])
def detach_from_suite(suite_id: int, task_ids: list[int]) -> int:
"""Detaches a set of task ids from a suite.
Parameters
----------
suite_id : int
OpenML id of the study
task_ids : list (int)
List of entities to unlink from the collection
Returns
-------
int
new size of the study (in terms of explicitly linked entities)
"""
return detach_from_study(suite_id, task_ids)
def detach_from_study(study_id: int, run_ids: list[int]) -> int:
"""Detaches a set of run ids from a study.
Parameters
----------
study_id : int
OpenML id of the study
run_ids : list (int)
List of entities to unlink from the collection
Returns
-------
int
new size of the study (in terms of explicitly linked entities)
"""
# Interestingly, there's no need to tell the server about the entity type, it knows by itself
uri = "study/%d/detach" % study_id
post_variables = {"ids": ",".join(str(x) for x in run_ids)} # type: openml._api_calls.DATA_TYPE
result_xml = openml._api_calls._perform_api_call(
call=uri,
request_method="post",
data=post_variables,
)
result = xmltodict.parse(result_xml)["oml:study_detach"]
return int(result["oml:linked_entities"])
def list_suites(
offset: int | None = None,
size: int | None = None,
status: str | None = None,
uploader: list[int] | None = None,
) -> pd.DataFrame:
"""
Return a list of all suites which are on OpenML.
Parameters
----------
offset : int, optional
The number of suites to skip, starting from the first.
size : int, optional
The maximum number of suites to show.
status : str, optional
Should be {active, in_preparation, deactivated, all}. By default active
suites are returned.
uploader : list (int), optional
Result filter. Will only return suites created by these users.
Returns
-------
datasets : dataframe
Every row is represented by a dictionary containing the following information:
- id
- alias (optional)
- name
- main_entity_type
- status
- creator
- creation_date
"""
listing_call = partial(
_list_studies,
main_entity_type="task",
status=status,
uploader=uploader,
)
batches = openml.utils._list_all(listing_call, limit=size, offset=offset)
if len(batches) == 0:
return pd.DataFrame()
return pd.concat(batches)
def list_studies(
offset: int | None = None,
size: int | None = None,
status: str | None = None,
uploader: list[str] | None = None,
benchmark_suite: int | None = None,
) -> pd.DataFrame:
"""
Return a list of all studies which are on OpenML.
Parameters
----------
offset : int, optional
The number of studies to skip, starting from the first.
size : int, optional
The maximum number of studies to show.
status : str, optional
Should be {active, in_preparation, deactivated, all}. By default active
studies are returned.
uploader : list (int), optional
Result filter. Will only return studies created by these users.
benchmark_suite : int, optional
Returns
-------
datasets : dataframe
Every dataset is represented by a dictionary containing
the following information:
- id
- alias (optional)
- name
- benchmark_suite (optional)
- status
- creator
- creation_date
If qualities are calculated for the dataset, some of
these are also returned.
"""
listing_call = partial(
_list_studies,
main_entity_type="run",
status=status,
uploader=uploader,
benchmark_suite=benchmark_suite,
)
batches = openml.utils._list_all(listing_call, offset=offset, limit=size)
if len(batches) == 0:
return pd.DataFrame()
return pd.concat(batches)
def _list_studies(limit: int, offset: int, **kwargs: Any) -> pd.DataFrame:
"""Perform api call to return a list of studies.
Parameters
----------
limit: int
The maximum number of studies to return.
offset: int
The number of studies to skip, starting from the first.
kwargs : dict, optional
Legal filter operators (keys in the dict):
status, main_entity_type, uploader, benchmark_suite
Returns
-------
studies : dataframe
"""
api_call = "study/list"
if limit is not None:
api_call += f"/limit/{limit}"
if offset is not None:
api_call += f"/offset/{offset}"
if kwargs is not None:
for operator, value in kwargs.items():
if value is not None:
api_call += f"/{operator}/{value}"
return __list_studies(api_call=api_call)
def __list_studies(api_call: str) -> pd.DataFrame:
"""Retrieves the list of OpenML studies and
returns it in a dictionary or a Pandas DataFrame.
Parameters
----------
api_call : str
The API call for retrieving the list of OpenML studies.
Returns
-------
pd.DataFrame
A Pandas DataFrame of OpenML studies
"""
xml_string = openml._api_calls._perform_api_call(api_call, "get")
study_dict = xmltodict.parse(xml_string, force_list=("oml:study",))
# Minimalistic check if the XML is useful
assert isinstance(study_dict["oml:study_list"]["oml:study"], list), type(
study_dict["oml:study_list"],
)
assert study_dict["oml:study_list"]["@xmlns:oml"] == "http://openml.org/openml", study_dict[
"oml:study_list"
]["@xmlns:oml"]
studies = {}
for study_ in study_dict["oml:study_list"]["oml:study"]:
# maps from xml name to a tuple of (dict name, casting fn)
expected_fields = {
"oml:id": ("id", int),
"oml:alias": ("alias", str),
"oml:main_entity_type": ("main_entity_type", str),
"oml:benchmark_suite": ("benchmark_suite", int),
"oml:name": ("name", str),
"oml:status": ("status", str),
"oml:creation_date": ("creation_date", str),
"oml:creator": ("creator", int),
}
study_id = int(study_["oml:id"])
current_study = {}
for oml_field_name, (real_field_name, cast_fn) in expected_fields.items():
if oml_field_name in study_:
current_study[real_field_name] = cast_fn(study_[oml_field_name])
current_study["id"] = int(current_study["id"])
studies[study_id] = current_study
return pd.DataFrame.from_dict(studies, orient="index")