diff options
Diffstat (limited to 'python/pyspark/ml/clustering.py')
-rw-r--r-- | python/pyspark/ml/clustering.py | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/python/pyspark/ml/clustering.py b/python/pyspark/ml/clustering.py index 64c4bf1b92..05aa2dfe74 100644 --- a/python/pyspark/ml/clustering.py +++ b/python/pyspark/ml/clustering.py @@ -92,7 +92,8 @@ class KMeans(JavaEstimator, HasFeaturesCol, HasPredictionCol, HasMaxIter, HasTol initMode = Param(Params._dummy(), "initMode", "the initialization algorithm. This can be either \"random\" to " + "choose random points as initial cluster centers, or \"k-means||\" " + - "to use a parallel variant of k-means++", TypeConverters.toString) + "to use a parallel variant of k-means++", + typeConverter=TypeConverters.toString) initSteps = Param(Params._dummy(), "initSteps", "steps for k-means initialization mode", typeConverter=TypeConverters.toInt) |