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
|
#
# 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 pyspark.ml.util import keyword_only
from pyspark.ml.wrapper import JavaEstimator, JavaModel
from pyspark.ml.param.shared import *
from pyspark.mllib.common import inherit_doc
__all__ = ['DecisionTreeRegressor', 'DecisionTreeRegressionModel', 'GBTRegressor',
'GBTRegressionModel', 'LinearRegression', 'LinearRegressionModel',
'RandomForestRegressor', 'RandomForestRegressionModel']
@inherit_doc
class LinearRegression(JavaEstimator, HasFeaturesCol, HasLabelCol, HasPredictionCol, HasMaxIter,
HasRegParam, HasTol, HasElasticNetParam, HasFitIntercept,
HasStandardization):
"""
Linear regression.
The learning objective is to minimize the squared error, with regularization.
The specific squared error loss function used is: L = 1/2n ||A weights - y||^2^
This support multiple types of regularization:
- none (a.k.a. ordinary least squares)
- L2 (ridge regression)
- L1 (Lasso)
- L2 + L1 (elastic net)
>>> from pyspark.mllib.linalg import Vectors
>>> df = sqlContext.createDataFrame([
... (1.0, Vectors.dense(1.0)),
... (0.0, Vectors.sparse(1, [], []))], ["label", "features"])
>>> lr = LinearRegression(maxIter=5, regParam=0.0)
>>> model = lr.fit(df)
>>> test0 = sqlContext.createDataFrame([(Vectors.dense(-1.0),)], ["features"])
>>> model.transform(test0).head().prediction
-1.0
>>> model.weights
DenseVector([1.0])
>>> model.intercept
0.0
>>> test1 = sqlContext.createDataFrame([(Vectors.sparse(1, [0], [1.0]),)], ["features"])
>>> model.transform(test1).head().prediction
1.0
>>> lr.setParams("vector")
Traceback (most recent call last):
...
TypeError: Method setParams forces keyword arguments.
"""
@keyword_only
def __init__(self, featuresCol="features", labelCol="label", predictionCol="prediction",
maxIter=100, regParam=0.0, elasticNetParam=0.0, tol=1e-6, fitIntercept=True,
standardization=True):
"""
__init__(self, featuresCol="features", labelCol="label", predictionCol="prediction", \
maxIter=100, regParam=0.0, elasticNetParam=0.0, tol=1e-6, fitIntercept=True, \
standardization=True)
"""
super(LinearRegression, self).__init__()
self._java_obj = self._new_java_obj(
"org.apache.spark.ml.regression.LinearRegression", self.uid)
self._setDefault(maxIter=100, regParam=0.0, tol=1e-6)
kwargs = self.__init__._input_kwargs
self.setParams(**kwargs)
@keyword_only
def setParams(self, featuresCol="features", labelCol="label", predictionCol="prediction",
maxIter=100, regParam=0.0, elasticNetParam=0.0, tol=1e-6, fitIntercept=True,
standardization=True):
"""
setParams(self, featuresCol="features", labelCol="label", predictionCol="prediction", \
maxIter=100, regParam=0.0, elasticNetParam=0.0, tol=1e-6, fitIntercept=True, \
standardization=True)
Sets params for linear regression.
"""
kwargs = self.setParams._input_kwargs
return self._set(**kwargs)
def _create_model(self, java_model):
return LinearRegressionModel(java_model)
class LinearRegressionModel(JavaModel):
"""
Model fitted by LinearRegression.
"""
@property
def weights(self):
"""
Model weights.
"""
return self._call_java("weights")
@property
def intercept(self):
"""
Model intercept.
"""
return self._call_java("intercept")
class TreeRegressorParams(object):
"""
Private class to track supported impurity measures.
"""
supportedImpurities = ["variance"]
class RandomForestParams(object):
"""
Private class to track supported random forest parameters.
"""
supportedFeatureSubsetStrategies = ["auto", "all", "onethird", "sqrt", "log2"]
class GBTParams(object):
"""
Private class to track supported GBT params.
"""
supportedLossTypes = ["squared", "absolute"]
@inherit_doc
class DecisionTreeRegressor(JavaEstimator, HasFeaturesCol, HasLabelCol, HasPredictionCol,
DecisionTreeParams, HasCheckpointInterval):
"""
`http://en.wikipedia.org/wiki/Decision_tree_learning Decision tree`
learning algorithm for regression.
It supports both continuous and categorical features.
>>> from pyspark.mllib.linalg import Vectors
>>> df = sqlContext.createDataFrame([
... (1.0, Vectors.dense(1.0)),
... (0.0, Vectors.sparse(1, [], []))], ["label", "features"])
>>> dt = DecisionTreeRegressor(maxDepth=2)
>>> model = dt.fit(df)
>>> model.depth
1
>>> model.numNodes
3
>>> test0 = sqlContext.createDataFrame([(Vectors.dense(-1.0),)], ["features"])
>>> model.transform(test0).head().prediction
0.0
>>> test1 = sqlContext.createDataFrame([(Vectors.sparse(1, [0], [1.0]),)], ["features"])
>>> model.transform(test1).head().prediction
1.0
"""
# a placeholder to make it appear in the generated doc
impurity = Param(Params._dummy(), "impurity",
"Criterion used for information gain calculation (case-insensitive). " +
"Supported options: " + ", ".join(TreeRegressorParams.supportedImpurities))
@keyword_only
def __init__(self, featuresCol="features", labelCol="label", predictionCol="prediction",
maxDepth=5, maxBins=32, minInstancesPerNode=1, minInfoGain=0.0,
maxMemoryInMB=256, cacheNodeIds=False, checkpointInterval=10, impurity="variance"):
"""
__init__(self, featuresCol="features", labelCol="label", predictionCol="prediction", \
maxDepth=5, maxBins=32, minInstancesPerNode=1, minInfoGain=0.0, \
maxMemoryInMB=256, cacheNodeIds=False, checkpointInterval=10, impurity="variance")
"""
super(DecisionTreeRegressor, self).__init__()
self._java_obj = self._new_java_obj(
"org.apache.spark.ml.regression.DecisionTreeRegressor", self.uid)
#: param for Criterion used for information gain calculation (case-insensitive).
self.impurity = \
Param(self, "impurity",
"Criterion used for information gain calculation (case-insensitive). " +
"Supported options: " + ", ".join(TreeRegressorParams.supportedImpurities))
self._setDefault(maxDepth=5, maxBins=32, minInstancesPerNode=1, minInfoGain=0.0,
maxMemoryInMB=256, cacheNodeIds=False, checkpointInterval=10,
impurity="variance")
kwargs = self.__init__._input_kwargs
self.setParams(**kwargs)
@keyword_only
def setParams(self, featuresCol="features", labelCol="label", predictionCol="prediction",
maxDepth=5, maxBins=32, minInstancesPerNode=1, minInfoGain=0.0,
maxMemoryInMB=256, cacheNodeIds=False, checkpointInterval=10,
impurity="variance"):
"""
setParams(self, featuresCol="features", labelCol="label", predictionCol="prediction", \
maxDepth=5, maxBins=32, minInstancesPerNode=1, minInfoGain=0.0, \
maxMemoryInMB=256, cacheNodeIds=False, checkpointInterval=10, impurity="variance")
Sets params for the DecisionTreeRegressor.
"""
kwargs = self.setParams._input_kwargs
return self._set(**kwargs)
def _create_model(self, java_model):
return DecisionTreeRegressionModel(java_model)
def setImpurity(self, value):
"""
Sets the value of :py:attr:`impurity`.
"""
self._paramMap[self.impurity] = value
return self
def getImpurity(self):
"""
Gets the value of impurity or its default value.
"""
return self.getOrDefault(self.impurity)
@inherit_doc
class DecisionTreeModel(JavaModel):
@property
def numNodes(self):
"""Return number of nodes of the decision tree."""
return self._call_java("numNodes")
@property
def depth(self):
"""Return depth of the decision tree."""
return self._call_java("depth")
def __repr__(self):
return self._call_java("toString")
@inherit_doc
class TreeEnsembleModels(JavaModel):
@property
def treeWeights(self):
"""Return the weights for each tree"""
return list(self._call_java("javaTreeWeights"))
def __repr__(self):
return self._call_java("toString")
@inherit_doc
class DecisionTreeRegressionModel(DecisionTreeModel):
"""
Model fitted by DecisionTreeRegressor.
"""
@inherit_doc
class RandomForestRegressor(JavaEstimator, HasFeaturesCol, HasLabelCol, HasPredictionCol, HasSeed,
DecisionTreeParams, HasCheckpointInterval):
"""
`http://en.wikipedia.org/wiki/Random_forest Random Forest`
learning algorithm for regression.
It supports both continuous and categorical features.
>>> from numpy import allclose
>>> from pyspark.mllib.linalg import Vectors
>>> df = sqlContext.createDataFrame([
... (1.0, Vectors.dense(1.0)),
... (0.0, Vectors.sparse(1, [], []))], ["label", "features"])
>>> rf = RandomForestRegressor(numTrees=2, maxDepth=2, seed=42)
>>> model = rf.fit(df)
>>> allclose(model.treeWeights, [1.0, 1.0])
True
>>> test0 = sqlContext.createDataFrame([(Vectors.dense(-1.0),)], ["features"])
>>> model.transform(test0).head().prediction
0.0
>>> test1 = sqlContext.createDataFrame([(Vectors.sparse(1, [0], [1.0]),)], ["features"])
>>> model.transform(test1).head().prediction
0.5
"""
# a placeholder to make it appear in the generated doc
impurity = Param(Params._dummy(), "impurity",
"Criterion used for information gain calculation (case-insensitive). " +
"Supported options: " + ", ".join(TreeRegressorParams.supportedImpurities))
subsamplingRate = Param(Params._dummy(), "subsamplingRate",
"Fraction of the training data used for learning each decision tree, " +
"in range (0, 1].")
numTrees = Param(Params._dummy(), "numTrees", "Number of trees to train (>= 1)")
featureSubsetStrategy = \
Param(Params._dummy(), "featureSubsetStrategy",
"The number of features to consider for splits at each tree node. Supported " +
"options: " + ", ".join(RandomForestParams.supportedFeatureSubsetStrategies))
@keyword_only
def __init__(self, featuresCol="features", labelCol="label", predictionCol="prediction",
maxDepth=5, maxBins=32, minInstancesPerNode=1, minInfoGain=0.0,
maxMemoryInMB=256, cacheNodeIds=False, checkpointInterval=10, impurity="variance",
numTrees=20, featureSubsetStrategy="auto", seed=None):
"""
__init__(self, featuresCol="features", labelCol="label", predictionCol="prediction", \
maxDepth=5, maxBins=32, minInstancesPerNode=1, minInfoGain=0.0, \
maxMemoryInMB=256, cacheNodeIds=False, checkpointInterval=10, \
impurity="variance", numTrees=20, \
featureSubsetStrategy="auto", seed=None)
"""
super(RandomForestRegressor, self).__init__()
self._java_obj = self._new_java_obj(
"org.apache.spark.ml.regression.RandomForestRegressor", self.uid)
#: param for Criterion used for information gain calculation (case-insensitive).
self.impurity = \
Param(self, "impurity",
"Criterion used for information gain calculation (case-insensitive). " +
"Supported options: " + ", ".join(TreeRegressorParams.supportedImpurities))
#: param for Fraction of the training data used for learning each decision tree,
# in range (0, 1]
self.subsamplingRate = Param(self, "subsamplingRate",
"Fraction of the training data used for learning each " +
"decision tree, in range (0, 1].")
#: param for Number of trees to train (>= 1)
self.numTrees = Param(self, "numTrees", "Number of trees to train (>= 1)")
#: param for The number of features to consider for splits at each tree node
self.featureSubsetStrategy = \
Param(self, "featureSubsetStrategy",
"The number of features to consider for splits at each tree node. Supported " +
"options: " + ", ".join(RandomForestParams.supportedFeatureSubsetStrategies))
self._setDefault(maxDepth=5, maxBins=32, minInstancesPerNode=1, minInfoGain=0.0,
maxMemoryInMB=256, cacheNodeIds=False, checkpointInterval=10, seed=None,
impurity="variance", numTrees=20, featureSubsetStrategy="auto")
kwargs = self.__init__._input_kwargs
self.setParams(**kwargs)
@keyword_only
def setParams(self, featuresCol="features", labelCol="label", predictionCol="prediction",
maxDepth=5, maxBins=32, minInstancesPerNode=1, minInfoGain=0.0,
maxMemoryInMB=256, cacheNodeIds=False, checkpointInterval=10, seed=None,
impurity="variance", numTrees=20, featureSubsetStrategy="auto"):
"""
setParams(self, featuresCol="features", labelCol="label", predictionCol="prediction", \
maxDepth=5, maxBins=32, minInstancesPerNode=1, minInfoGain=0.0, \
maxMemoryInMB=256, cacheNodeIds=False, checkpointInterval=10, seed=None, \
impurity="variance", numTrees=20, featureSubsetStrategy="auto")
Sets params for linear regression.
"""
kwargs = self.setParams._input_kwargs
return self._set(**kwargs)
def _create_model(self, java_model):
return RandomForestRegressionModel(java_model)
def setImpurity(self, value):
"""
Sets the value of :py:attr:`impurity`.
"""
self._paramMap[self.impurity] = value
return self
def getImpurity(self):
"""
Gets the value of impurity or its default value.
"""
return self.getOrDefault(self.impurity)
def setSubsamplingRate(self, value):
"""
Sets the value of :py:attr:`subsamplingRate`.
"""
self._paramMap[self.subsamplingRate] = value
return self
def getSubsamplingRate(self):
"""
Gets the value of subsamplingRate or its default value.
"""
return self.getOrDefault(self.subsamplingRate)
def setNumTrees(self, value):
"""
Sets the value of :py:attr:`numTrees`.
"""
self._paramMap[self.numTrees] = value
return self
def getNumTrees(self):
"""
Gets the value of numTrees or its default value.
"""
return self.getOrDefault(self.numTrees)
def setFeatureSubsetStrategy(self, value):
"""
Sets the value of :py:attr:`featureSubsetStrategy`.
"""
self._paramMap[self.featureSubsetStrategy] = value
return self
def getFeatureSubsetStrategy(self):
"""
Gets the value of featureSubsetStrategy or its default value.
"""
return self.getOrDefault(self.featureSubsetStrategy)
class RandomForestRegressionModel(TreeEnsembleModels):
"""
Model fitted by RandomForestRegressor.
"""
@inherit_doc
class GBTRegressor(JavaEstimator, HasFeaturesCol, HasLabelCol, HasPredictionCol, HasMaxIter,
DecisionTreeParams, HasCheckpointInterval):
"""
`http://en.wikipedia.org/wiki/Gradient_boosting Gradient-Boosted Trees (GBTs)`
learning algorithm for regression.
It supports both continuous and categorical features.
>>> from numpy import allclose
>>> from pyspark.mllib.linalg import Vectors
>>> df = sqlContext.createDataFrame([
... (1.0, Vectors.dense(1.0)),
... (0.0, Vectors.sparse(1, [], []))], ["label", "features"])
>>> gbt = GBTRegressor(maxIter=5, maxDepth=2)
>>> model = gbt.fit(df)
>>> allclose(model.treeWeights, [1.0, 0.1, 0.1, 0.1, 0.1])
True
>>> test0 = sqlContext.createDataFrame([(Vectors.dense(-1.0),)], ["features"])
>>> model.transform(test0).head().prediction
0.0
>>> test1 = sqlContext.createDataFrame([(Vectors.sparse(1, [0], [1.0]),)], ["features"])
>>> model.transform(test1).head().prediction
1.0
"""
# a placeholder to make it appear in the generated doc
lossType = Param(Params._dummy(), "lossType",
"Loss function which GBT tries to minimize (case-insensitive). " +
"Supported options: " + ", ".join(GBTParams.supportedLossTypes))
subsamplingRate = Param(Params._dummy(), "subsamplingRate",
"Fraction of the training data used for learning each decision tree, " +
"in range (0, 1].")
stepSize = Param(Params._dummy(), "stepSize",
"Step size (a.k.a. learning rate) in interval (0, 1] for shrinking the " +
"contribution of each estimator")
@keyword_only
def __init__(self, featuresCol="features", labelCol="label", predictionCol="prediction",
maxDepth=5, maxBins=32, minInstancesPerNode=1, minInfoGain=0.0,
maxMemoryInMB=256, cacheNodeIds=False, checkpointInterval=10, lossType="squared",
maxIter=20, stepSize=0.1):
"""
__init__(self, featuresCol="features", labelCol="label", predictionCol="prediction", \
maxDepth=5, maxBins=32, minInstancesPerNode=1, minInfoGain=0.0, \
maxMemoryInMB=256, cacheNodeIds=False, checkpointInterval=10, \
lossType="squared", maxIter=20, stepSize=0.1)
"""
super(GBTRegressor, self).__init__()
self._java_obj = self._new_java_obj("org.apache.spark.ml.regression.GBTRegressor", self.uid)
#: param for Loss function which GBT tries to minimize (case-insensitive).
self.lossType = Param(self, "lossType",
"Loss function which GBT tries to minimize (case-insensitive). " +
"Supported options: " + ", ".join(GBTParams.supportedLossTypes))
#: Fraction of the training data used for learning each decision tree, in range (0, 1].
self.subsamplingRate = Param(self, "subsamplingRate",
"Fraction of the training data used for learning each " +
"decision tree, in range (0, 1].")
#: Step size (a.k.a. learning rate) in interval (0, 1] for shrinking the contribution of
# each estimator
self.stepSize = Param(self, "stepSize",
"Step size (a.k.a. learning rate) in interval (0, 1] for shrinking " +
"the contribution of each estimator")
self._setDefault(maxDepth=5, maxBins=32, minInstancesPerNode=1, minInfoGain=0.0,
maxMemoryInMB=256, cacheNodeIds=False, checkpointInterval=10,
lossType="squared", maxIter=20, stepSize=0.1)
kwargs = self.__init__._input_kwargs
self.setParams(**kwargs)
@keyword_only
def setParams(self, featuresCol="features", labelCol="label", predictionCol="prediction",
maxDepth=5, maxBins=32, minInstancesPerNode=1, minInfoGain=0.0,
maxMemoryInMB=256, cacheNodeIds=False, checkpointInterval=10,
lossType="squared", maxIter=20, stepSize=0.1):
"""
setParams(self, featuresCol="features", labelCol="label", predictionCol="prediction", \
maxDepth=5, maxBins=32, minInstancesPerNode=1, minInfoGain=0.0, \
maxMemoryInMB=256, cacheNodeIds=False, checkpointInterval=10, \
lossType="squared", maxIter=20, stepSize=0.1)
Sets params for Gradient Boosted Tree Regression.
"""
kwargs = self.setParams._input_kwargs
return self._set(**kwargs)
def _create_model(self, java_model):
return GBTRegressionModel(java_model)
def setLossType(self, value):
"""
Sets the value of :py:attr:`lossType`.
"""
self._paramMap[self.lossType] = value
return self
def getLossType(self):
"""
Gets the value of lossType or its default value.
"""
return self.getOrDefault(self.lossType)
def setSubsamplingRate(self, value):
"""
Sets the value of :py:attr:`subsamplingRate`.
"""
self._paramMap[self.subsamplingRate] = value
return self
def getSubsamplingRate(self):
"""
Gets the value of subsamplingRate or its default value.
"""
return self.getOrDefault(self.subsamplingRate)
def setStepSize(self, value):
"""
Sets the value of :py:attr:`stepSize`.
"""
self._paramMap[self.stepSize] = value
return self
def getStepSize(self):
"""
Gets the value of stepSize or its default value.
"""
return self.getOrDefault(self.stepSize)
class GBTRegressionModel(TreeEnsembleModels):
"""
Model fitted by GBTRegressor.
"""
if __name__ == "__main__":
import doctest
from pyspark.context import SparkContext
from pyspark.sql import SQLContext
globs = globals().copy()
# The small batch size here ensures that we see multiple batches,
# even in these small test examples:
sc = SparkContext("local[2]", "ml.regression tests")
sqlContext = SQLContext(sc)
globs['sc'] = sc
globs['sqlContext'] = sqlContext
(failure_count, test_count) = doctest.testmod(globs=globs, optionflags=doctest.ELLIPSIS)
sc.stop()
if failure_count:
exit(-1)
|