-
Notifications
You must be signed in to change notification settings - Fork 28k
/
base.py
414 lines (342 loc) · 11.8 KB
/
base.py
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
#
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You 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
#
# http://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.
#
from abc import ABCMeta, abstractmethod
import copy
import threading
from typing import (
Any,
Callable,
Generic,
Iterator,
List,
Optional,
Sequence,
Tuple,
TypeVar,
Union,
cast,
overload,
TYPE_CHECKING,
)
from pyspark import since
from pyspark.ml.param import P
from pyspark.ml.common import inherit_doc
from pyspark.ml.param.shared import (
HasInputCol,
HasOutputCol,
HasLabelCol,
HasFeaturesCol,
HasPredictionCol,
Params,
)
from pyspark.sql.dataframe import DataFrame
from pyspark.sql.functions import udf
from pyspark.sql.types import DataType, StructField, StructType
if TYPE_CHECKING:
from pyspark.ml._typing import ParamMap
T = TypeVar("T")
M = TypeVar("M", bound="Transformer")
class _FitMultipleIterator(Generic[M]):
"""
Used by default implementation of Estimator.fitMultiple to produce models in a thread safe
iterator. This class handles the simple case of fitMultiple where each param map should be
fit independently.
Parameters
----------
fitSingleModel : function
Callable[[int], Transformer] which fits an estimator to a dataset.
`fitSingleModel` may be called up to `numModels` times, with a unique index each time.
Each call to `fitSingleModel` with an index should return the Model associated with
that index.
numModel : int
Number of models this iterator should produce.
Notes
-----
See :py:meth:`Estimator.fitMultiple` for more info.
"""
def __init__(self, fitSingleModel: Callable[[int], M], numModels: int):
""" """
self.fitSingleModel = fitSingleModel
self.numModel = numModels
self.counter = 0
self.lock = threading.Lock()
def __iter__(self) -> Iterator[Tuple[int, M]]:
return self
def __next__(self) -> Tuple[int, M]:
with self.lock:
index = self.counter
if index >= self.numModel:
raise StopIteration("No models remaining.")
self.counter += 1
return index, self.fitSingleModel(index)
def next(self) -> Tuple[int, M]:
"""For python2 compatibility."""
return self.__next__()
@inherit_doc
class Estimator(Params, Generic[M], metaclass=ABCMeta):
"""
Abstract class for estimators that fit models to data.
.. versionadded:: 1.3.0
"""
@abstractmethod
def _fit(self, dataset: DataFrame) -> M:
"""
Fits a model to the input dataset. This is called by the default implementation of fit.
Parameters
----------
dataset : :py:class:`pyspark.sql.DataFrame`
input dataset
Returns
-------
:class:`Transformer`
fitted model
"""
raise NotImplementedError()
def fitMultiple(
self, dataset: DataFrame, paramMaps: Sequence["ParamMap"]
) -> Iterator[Tuple[int, M]]:
"""
Fits a model to the input dataset for each param map in `paramMaps`.
.. versionadded:: 2.3.0
Parameters
----------
dataset : :py:class:`pyspark.sql.DataFrame`
input dataset.
paramMaps : :py:class:`collections.abc.Sequence`
A Sequence of param maps.
Returns
-------
:py:class:`_FitMultipleIterator`
A thread safe iterable which contains one model for each param map. Each
call to `next(modelIterator)` will return `(index, model)` where model was fit
using `paramMaps[index]`. `index` values may not be sequential.
"""
estimator = self.copy()
def fitSingleModel(index: int) -> M:
return estimator.fit(dataset, paramMaps[index])
return _FitMultipleIterator(fitSingleModel, len(paramMaps))
@overload
def fit(self, dataset: DataFrame, params: Optional["ParamMap"] = ...) -> M:
...
@overload
def fit(
self, dataset: DataFrame, params: Union[List["ParamMap"], Tuple["ParamMap"]]
) -> List[M]:
...
def fit(
self,
dataset: DataFrame,
params: Optional[Union["ParamMap", List["ParamMap"], Tuple["ParamMap"]]] = None,
) -> Union[M, List[M]]:
"""
Fits a model to the input dataset with optional parameters.
.. versionadded:: 1.3.0
Parameters
----------
dataset : :py:class:`pyspark.sql.DataFrame`
input dataset.
params : dict or list or tuple, optional
an optional param map that overrides embedded params. If a list/tuple of
param maps is given, this calls fit on each param map and returns a list of
models.
Returns
-------
:py:class:`Transformer` or a list of :py:class:`Transformer`
fitted model(s)
"""
if params is None:
params = dict()
if isinstance(params, (list, tuple)):
models: List[Optional[M]] = [None] * len(params)
for index, model in self.fitMultiple(dataset, params):
models[index] = model
return cast(List[M], models)
elif isinstance(params, dict):
if params:
return self.copy(params)._fit(dataset)
else:
return self._fit(dataset)
else:
raise TypeError(
"Params must be either a param map or a list/tuple of param maps, "
"but got %s." % type(params)
)
@inherit_doc
class Transformer(Params, metaclass=ABCMeta):
"""
Abstract class for transformers that transform one dataset into another.
.. versionadded:: 1.3.0
"""
@abstractmethod
def _transform(self, dataset: DataFrame) -> DataFrame:
"""
Transforms the input dataset.
Parameters
----------
dataset : :py:class:`pyspark.sql.DataFrame`
input dataset.
Returns
-------
:py:class:`pyspark.sql.DataFrame`
transformed dataset
"""
raise NotImplementedError()
def transform(self, dataset: DataFrame, params: Optional["ParamMap"] = None) -> DataFrame:
"""
Transforms the input dataset with optional parameters.
.. versionadded:: 1.3.0
Parameters
----------
dataset : :py:class:`pyspark.sql.DataFrame`
input dataset
params : dict, optional
an optional param map that overrides embedded params.
Returns
-------
:py:class:`pyspark.sql.DataFrame`
transformed dataset
"""
if params is None:
params = dict()
if isinstance(params, dict):
if params:
return self.copy(params)._transform(dataset)
else:
return self._transform(dataset)
else:
raise TypeError("Params must be a param map but got %s." % type(params))
@inherit_doc
class Model(Transformer, metaclass=ABCMeta):
"""
Abstract class for models that are fitted by estimators.
.. versionadded:: 1.4.0
"""
pass
@inherit_doc
class UnaryTransformer(HasInputCol, HasOutputCol, Transformer):
"""
Abstract class for transformers that take one input column, apply transformation,
and output the result as a new column.
.. versionadded:: 2.3.0
"""
def setInputCol(self: P, value: str) -> P:
"""
Sets the value of :py:attr:`inputCol`.
"""
return self._set(inputCol=value)
def setOutputCol(self: P, value: str) -> P:
"""
Sets the value of :py:attr:`outputCol`.
"""
return self._set(outputCol=value)
@abstractmethod
def createTransformFunc(self) -> Callable[..., Any]:
"""
Creates the transform function using the given param map. The input param map already takes
account of the embedded param map. So the param values should be determined
solely by the input param map.
"""
raise NotImplementedError()
@abstractmethod
def outputDataType(self) -> DataType:
"""
Returns the data type of the output column.
"""
raise NotImplementedError()
@abstractmethod
def validateInputType(self, inputType: DataType) -> None:
"""
Validates the input type. Throw an exception if it is invalid.
"""
raise NotImplementedError()
def transformSchema(self, schema: StructType) -> StructType:
inputType = schema[self.getInputCol()].dataType
self.validateInputType(inputType)
if self.getOutputCol() in schema.names:
raise ValueError("Output column %s already exists." % self.getOutputCol())
outputFields = copy.copy(schema.fields)
outputFields.append(StructField(self.getOutputCol(), self.outputDataType(), nullable=False))
return StructType(outputFields)
def _transform(self, dataset: DataFrame) -> DataFrame:
self.transformSchema(dataset.schema)
transformUDF = udf(self.createTransformFunc(), self.outputDataType())
transformedDataset = dataset.withColumn(
self.getOutputCol(), transformUDF(dataset[self.getInputCol()])
)
return transformedDataset
@inherit_doc
class _PredictorParams(HasLabelCol, HasFeaturesCol, HasPredictionCol):
"""
Params for :py:class:`Predictor` and :py:class:`PredictorModel`.
.. versionadded:: 3.0.0
"""
pass
@inherit_doc
class Predictor(Estimator[M], _PredictorParams, metaclass=ABCMeta):
"""
Estimator for prediction tasks (regression and classification).
"""
@since("3.0.0")
def setLabelCol(self: P, value: str) -> P:
"""
Sets the value of :py:attr:`labelCol`.
"""
return self._set(labelCol=value)
@since("3.0.0")
def setFeaturesCol(self: P, value: str) -> P:
"""
Sets the value of :py:attr:`featuresCol`.
"""
return self._set(featuresCol=value)
@since("3.0.0")
def setPredictionCol(self: P, value: str) -> P:
"""
Sets the value of :py:attr:`predictionCol`.
"""
return self._set(predictionCol=value)
@inherit_doc
class PredictionModel(Model, _PredictorParams, Generic[T], metaclass=ABCMeta):
"""
Model for prediction tasks (regression and classification).
"""
@since("3.0.0")
def setFeaturesCol(self: P, value: str) -> P:
"""
Sets the value of :py:attr:`featuresCol`.
"""
return self._set(featuresCol=value)
@since("3.0.0")
def setPredictionCol(self: P, value: str) -> P:
"""
Sets the value of :py:attr:`predictionCol`.
"""
return self._set(predictionCol=value)
@property # type: ignore[misc]
@abstractmethod
@since("2.1.0")
def numFeatures(self) -> int:
"""
Returns the number of features the model was trained on. If unknown, returns -1
"""
raise NotImplementedError()
@abstractmethod
@since("3.0.0")
def predict(self, value: T) -> float:
"""
Predict label for the given features.
"""
raise NotImplementedError()