[SPARK-5962] [MLlib] Python support for Power Iteration Clustering

[yanboliang]

Python support for Power Iteration Clustering
https://issues.apache.org/jira/browse/SPARK-5962

[AmplabJenkins]

Merged build triggered.

[AmplabJenkins]

Merged build started.

[SparkQA]

Test build #35696 has started for PR 6992 at commit 4be4423.

[SparkQA]

Test build #35696 has finished for PR 6992 at commit 4be4423.

• This patch passes all tests.
• This patch merges cleanly.
• This patch adds the following public classes (experimental):
  • class PowerIterationClusteringModel(JavaModelWrapper, JavaSaveable, JavaLoader):
  • class PowerIterationClustering(object):
  • class Assignment(namedtuple("Assignment", ["id", "cluster"])):

[AmplabJenkins]

Merged build finished. Test PASSed.

[yanboliang]

@jkbradley @mengxr

[mengxr]

getAssignments is only called by Python. So the return type should be RDD[Array[Any]]. This might work:

def getAssignments: RDD[Array[Any]] = {
  model.assignments.map(x => Array(x.id, x.cluster))
}

[AmplabJenkins]

Merged build triggered.

[AmplabJenkins]

Merged build started.

[SparkQA]

Test build #35928 has started for PR 6992 at commit 6b03d82.

[SparkQA]

Test build #35928 has finished for PR 6992 at commit 6b03d82.

• This patch passes all tests.
• This patch merges cleanly.
• This patch adds the following public classes (experimental):
  • class PowerIterationClusteringModel(JavaModelWrapper, JavaSaveable, JavaLoader):
  • class PowerIterationClustering(object):
  • class Assignment(namedtuple("Assignment", ["id", "cluster"])):

[AmplabJenkins]

Merged build finished. Test PASSed.

[mengxr]

LGTM. Merged into master. Thanks!

[jkbradley]

@yanboliang Could you please make a JIRA and PR for updating the user guide with the Python API? Thanks!

[yanboliang]

@jkbradley OK, I will do it.