diff options
author | lewuathe <lewuathe@me.com> | 2015-03-20 17:18:18 -0400 |
---|---|---|
committer | Xiangrui Meng <meng@databricks.com> | 2015-03-20 17:18:18 -0400 |
commit | 257cde7c363efb3317bfb5c13975cca9154894e2 (patch) | |
tree | 420d26c15072875caef30987209539debaf8888f /python/pyspark/mllib | |
parent | 11e025956be3818c00effef0d650734f8feeb436 (diff) | |
download | spark-257cde7c363efb3317bfb5c13975cca9154894e2.tar.gz spark-257cde7c363efb3317bfb5c13975cca9154894e2.tar.bz2 spark-257cde7c363efb3317bfb5c13975cca9154894e2.zip |
[SPARK-6421][MLLIB] _regression_train_wrapper does not test initialWeights correctly
Weight parameters must be initialized correctly even when numpy array is passed as initial weights.
Author: lewuathe <lewuathe@me.com>
Closes #5101 from Lewuathe/SPARK-6421 and squashes the following commits:
7795201 [lewuathe] Fix lint-python errors
21d4fe3 [lewuathe] Fix init logic of weights
Diffstat (limited to 'python/pyspark/mllib')
-rw-r--r-- | python/pyspark/mllib/regression.py | 3 | ||||
-rw-r--r-- | python/pyspark/mllib/tests.py | 7 |
2 files changed, 9 insertions, 1 deletions
diff --git a/python/pyspark/mllib/regression.py b/python/pyspark/mllib/regression.py index 015a786011..414a0ada80 100644 --- a/python/pyspark/mllib/regression.py +++ b/python/pyspark/mllib/regression.py @@ -163,7 +163,8 @@ def _regression_train_wrapper(train_func, modelClass, data, initial_weights): first = data.first() if not isinstance(first, LabeledPoint): raise ValueError("data should be an RDD of LabeledPoint, but got %s" % first) - initial_weights = initial_weights or [0.0] * len(data.first().features) + if initial_weights is None: + initial_weights = [0.0] * len(data.first().features) weights, intercept = train_func(data, _convert_to_vector(initial_weights)) return modelClass(weights, intercept) diff --git a/python/pyspark/mllib/tests.py b/python/pyspark/mllib/tests.py index 5328d99b69..155019638f 100644 --- a/python/pyspark/mllib/tests.py +++ b/python/pyspark/mllib/tests.py @@ -323,6 +323,13 @@ class ListTests(PySparkTestCase): self.assertTrue(gbt_model.predict(features[2]) <= 0) self.assertTrue(gbt_model.predict(features[3]) > 0) + try: + LinearRegressionWithSGD.train(rdd, initialWeights=array([1.0, 1.0])) + LassoWithSGD.train(rdd, initialWeights=array([1.0, 1.0])) + RidgeRegressionWithSGD.train(rdd, initialWeights=array([1.0, 1.0])) + except ValueError: + self.fail() + class StatTests(PySparkTestCase): # SPARK-4023 |