[SPARK-2850] [SPARK-2626] [mllib] MLlib stats examples + small fixes

Added examples for statistical summarization:
* Scala: StatisticalSummary.scala
** Tests: correlation, MultivariateOnlineSummarizer
* python: statistical_summary.py
** Tests: correlation (since MultivariateOnlineSummarizer has no Python API)

Added examples for random and sampled RDDs:
* Scala: RandomAndSampledRDDs.scala
* python: random_and_sampled_rdds.py
* Both test:
** RandomRDDGenerators.normalRDD, normalVectorRDD
** RDD.sample, takeSample, sampleByKey

Added sc.stop() to all examples.

CorrelationSuite.scala
* Added 1 test for RDDs with only 1 value

RowMatrix.scala
* numCols(): Added check for numRows = 0, with error message.
* computeCovariance(): Added check for numRows <= 1, with error message.

Python SparseVector (pyspark/mllib/linalg.py)
* Added toDense() function

python/run-tests script
* Added stat.py (doc test)

CC: mengxr dorx  Main changes were examples to show usage across APIs.

Author: Joseph K. Bradley <joseph.kurata.bradley@gmail.com>

Closes #1878 from jkbradley/mllib-stats-api-check and squashes the following commits:

ea5c047 [Joseph K. Bradley] Merge remote-tracking branch 'upstream/master' into mllib-stats-api-check
dafebe2 [Joseph K. Bradley] Bug fixes for examples SampledRDDs.scala and sampled_rdds.py: Check for division by 0 and for missing key in maps.
8d1e555 [Joseph K. Bradley] Merge remote-tracking branch 'upstream/master' into mllib-stats-api-check
60c72d9 [Joseph K. Bradley] Fixed stat.py doc test to work for Python versions printing nan or NaN.
b20d90a [Joseph K. Bradley] Merge remote-tracking branch 'upstream/master' into mllib-stats-api-check
4e5d15e [Joseph K. Bradley] Changed pyspark/mllib/stat.py doc tests to use NaN instead of nan.
32173b7 [Joseph K. Bradley] Stats examples update.
c8c20dc [Joseph K. Bradley] Merge remote-tracking branch 'upstream/master' into mllib-stats-api-check
cf70b07 [Joseph K. Bradley] Merge remote-tracking branch 'upstream/master' into mllib-stats-api-check
0b7cec3 [Joseph K. Bradley] Small updates based on code review.  Renamed statistical_summary.py to correlations.py
ab48f6e [Joseph K. Bradley] RowMatrix.scala * numCols(): Added check for numRows = 0, with error message. * computeCovariance(): Added check for numRows <= 1, with error message.
65e4ebc [Joseph K. Bradley] Merge remote-tracking branch 'upstream/master' into mllib-stats-api-check
8195c78 [Joseph K. Bradley] Added examples for random and sampled RDDs: * Scala: RandomAndSampledRDDs.scala * python: random_and_sampled_rdds.py * Both test: ** RandomRDDGenerators.normalRDD, normalVectorRDD ** RDD.sample, takeSample, sampleByKey
064985b [Joseph K. Bradley] Merge remote-tracking branch 'upstream/master' into mllib-stats-api-check
ee918e9 [Joseph K. Bradley] Added examples for statistical summarization: * Scala: StatisticalSummary.scala ** Tests: correlation, MultivariateOnlineSummarizer * python: statistical_summary.py ** Tests: correlation (since MultivariateOnlineSummarizer has no Python API)

(cherry picked from commit c8b16ca)
Signed-off-by: Xiangrui Meng <meng@databricks.com>

examples/src/main/python/als.py

@@ -97,3 +97,5 @@ def update(i, vec, mat, ratings):
         error = rmse(R, ms, us)
         print "Iteration %d:" % i
         print "\nRMSE: %5.4f\n" % error
+
+    sc.stop()

examples/src/main/python/cassandra_inputformat.py

@@ -77,3 +77,5 @@
     output = cass_rdd.collect()
     for (k, v) in output:
         print (k, v)
+
+    sc.stop()

examples/src/main/python/cassandra_outputformat.py

@@ -81,3 +81,5 @@
         conf=conf,
         keyConverter="org.apache.spark.examples.pythonconverters.ToCassandraCQLKeyConverter",
         valueConverter="org.apache.spark.examples.pythonconverters.ToCassandraCQLValueConverter")
+
+    sc.stop()

examples/src/main/python/hbase_inputformat.py

@@ -71,3 +71,5 @@
     output = hbase_rdd.collect()
     for (k, v) in output:
         print (k, v)
+
+    sc.stop()

examples/src/main/python/hbase_outputformat.py

@@ -63,3 +63,5 @@
         conf=conf,
         keyConverter="org.apache.spark.examples.pythonconverters.StringToImmutableBytesWritableConverter",
         valueConverter="org.apache.spark.examples.pythonconverters.StringListToPutConverter")
+
+    sc.stop()

examples/src/main/python/kmeans.py

@@ -77,3 +77,5 @@ def closestPoint(p, centers):
             kPoints[x] = y
 
     print "Final centers: " + str(kPoints)
+
+    sc.stop()

examples/src/main/python/logistic_regression.py

@@ -80,3 +80,5 @@ def add(x, y):
         w -= points.map(lambda m: gradient(m, w)).reduce(add)
 
     print "Final w: " + str(w)
+
+    sc.stop()

examples/src/main/python/mllib/correlations.py

@@ -0,0 +1,60 @@
+#
+# 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.
+#
+
+"""
+Correlations using MLlib.
+"""
+
+import sys
+
+from pyspark import SparkContext
+from pyspark.mllib.regression import LabeledPoint
+from pyspark.mllib.stat import Statistics
+from pyspark.mllib.util import MLUtils
+
+
+if __name__ == "__main__":
+    if len(sys.argv) not in [1,2]:
+        print >> sys.stderr, "Usage: correlations (<file>)"
+        exit(-1)
+    sc = SparkContext(appName="PythonCorrelations")
+    if len(sys.argv) == 2:
+        filepath = sys.argv[1]
+    else:
+        filepath = 'data/mllib/sample_linear_regression_data.txt'
+    corrType = 'pearson'
+
+    points = MLUtils.loadLibSVMFile(sc, filepath)\
+        .map(lambda lp: LabeledPoint(lp.label, lp.features.toArray()))
+
+    print
+    print 'Summary of data file: ' + filepath
+    print '%d data points' % points.count()
+
+    # Statistics (correlations)
+    print
+    print 'Correlation (%s) between label and each feature' % corrType
+    print 'Feature\tCorrelation'
+    numFeatures = points.take(1)[0].features.size
+    labelRDD = points.map(lambda lp: lp.label)
+    for i in range(numFeatures):
+        featureRDD = points.map(lambda lp: lp.features[i])
+        corr = Statistics.corr(labelRDD, featureRDD, corrType)
+        print '%d\t%g' % (i, corr)
+    print
+
+    sc.stop()

examples/src/main/python/mllib/decision_tree_runner.py

@@ -17,6 +17,8 @@
 
 """
 Decision tree classification and regression using MLlib.
+
+This example requires NumPy (http://www.numpy.org/).
 """
 
 import numpy, os, sys
@@ -117,6 +119,7 @@ def usage():
     if len(sys.argv) == 2:
         dataPath = sys.argv[1]
     if not os.path.isfile(dataPath):
+        sc.stop()
         usage()
     points = MLUtils.loadLibSVMFile(sc, dataPath)
 
@@ -133,3 +136,5 @@ def usage():
     print "  Model depth: %d\n" % model.depth()
     print "  Training accuracy: %g\n" % getAccuracy(model, reindexedData)
     print model
+
+    sc.stop()

examples/src/main/python/mllib/kmeans.py

@@ -42,3 +42,4 @@ def parseVector(line):
     k = int(sys.argv[2])
     model = KMeans.train(data, k)
     print "Final centers: " + str(model.clusterCenters)
+    sc.stop()

examples/src/main/python/mllib/logistic_regression.py

@@ -50,3 +50,4 @@ def parsePoint(line):
     model = LogisticRegressionWithSGD.train(points, iterations)
     print "Final weights: " + str(model.weights)
     print "Final intercept: " + str(model.intercept)
+    sc.stop()

examples/src/main/python/mllib/random_rdd_generation.py

@@ -0,0 +1,55 @@
+#
+# 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.
+#
+
+"""
+Randomly generated RDDs.
+"""
+
+import sys
+
+from pyspark import SparkContext
+from pyspark.mllib.random import RandomRDDs
+
+
+if __name__ == "__main__":
+    if len(sys.argv) not in [1, 2]:
+        print >> sys.stderr, "Usage: random_rdd_generation"
+        exit(-1)
+
+    sc = SparkContext(appName="PythonRandomRDDGeneration")
+
+    numExamples = 10000 # number of examples to generate
+    fraction = 0.1 # fraction of data to sample
+
+    # Example: RandomRDDs.normalRDD
+    normalRDD = RandomRDDs.normalRDD(sc, numExamples)
+    print 'Generated RDD of %d examples sampled from the standard normal distribution'\
+        % normalRDD.count()
+    print '  First 5 samples:'
+    for sample in normalRDD.take(5):
+        print '    ' + str(sample)
+    print
+
+    # Example: RandomRDDs.normalVectorRDD
+    normalVectorRDD = RandomRDDs.normalVectorRDD(sc, numRows = numExamples, numCols = 2)
+    print 'Generated RDD of %d examples of length-2 vectors.' % normalVectorRDD.count()
+    print '  First 5 samples:'
+    for sample in normalVectorRDD.take(5):
+        print '    ' + str(sample)
+    print
+
+    sc.stop()

examples/src/main/python/mllib/sampled_rdds.py

@@ -0,0 +1,86 @@
+#
+# 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.
+#
+
+"""
+Randomly sampled RDDs.
+"""
+
+import sys
+
+from pyspark import SparkContext
+from pyspark.mllib.util import MLUtils
+
+
+if __name__ == "__main__":
+    if len(sys.argv) not in [1, 2]:
+        print >> sys.stderr, "Usage: sampled_rdds <libsvm data file>"
+        exit(-1)
+    if len(sys.argv) == 2:
+        datapath = sys.argv[1]
+    else:
+        datapath = 'data/mllib/sample_binary_classification_data.txt'
+
+    sc = SparkContext(appName="PythonSampledRDDs")
+
+    fraction = 0.1 # fraction of data to sample
+
+    examples = MLUtils.loadLibSVMFile(sc, datapath)
+    numExamples = examples.count()
+    if numExamples == 0:
+        print >> sys.stderr, "Error: Data file had no samples to load."
+        exit(1)
+    print 'Loaded data with %d examples from file: %s' % (numExamples, datapath)
+
+    # Example: RDD.sample() and RDD.takeSample()
+    expectedSampleSize = int(numExamples * fraction)
+    print 'Sampling RDD using fraction %g.  Expected sample size = %d.' \
+        % (fraction, expectedSampleSize)
+    sampledRDD = examples.sample(withReplacement = True, fraction = fraction)
+    print '  RDD.sample(): sample has %d examples' % sampledRDD.count()
+    sampledArray = examples.takeSample(withReplacement = True, num = expectedSampleSize)
+    print '  RDD.takeSample(): sample has %d examples' % len(sampledArray)
+
+    print
+
+    # Example: RDD.sampleByKey()
+    keyedRDD = examples.map(lambda lp: (int(lp.label), lp.features))
+    print '  Keyed data using label (Int) as key ==> Orig'
+    #  Count examples per label in original data.
+    keyCountsA = keyedRDD.countByKey()
+
+    #  Subsample, and count examples per label in sampled data.
+    fractions = {}
+    for k in keyCountsA.keys():
+        fractions[k] = fraction
+    sampledByKeyRDD = keyedRDD.sampleByKey(withReplacement = True, fractions = fractions)
+    keyCountsB = sampledByKeyRDD.countByKey()
+    sizeB = sum(keyCountsB.values())
+    print '  Sampled %d examples using approximate stratified sampling (by label). ==> Sample' \
+        % sizeB
+
+    #  Compare samples
+    print '   \tFractions of examples with key'
+    print 'Key\tOrig\tSample'
+    for k in sorted(keyCountsA.keys()):
+        fracA = keyCountsA[k] / float(numExamples)
+        if sizeB != 0:
+            fracB = keyCountsB.get(k, 0) / float(sizeB)
+        else:
+            fracB = 0
+        print '%d\t%g\t%g' % (k, fracA, fracB)
+
+    sc.stop()

examples/src/main/python/pagerank.py

@@ -68,3 +68,5 @@ def parseNeighbors(urls):
     # Collects all URL ranks and dump them to console.
     for (link, rank) in ranks.collect():
         print "%s has rank: %s." % (link, rank)
+
+    sc.stop()

examples/src/main/python/pi.py

@@ -37,3 +37,5 @@ def f(_):
 
     count = sc.parallelize(xrange(1, n+1), slices).map(f).reduce(add)
     print "Pi is roughly %f" % (4.0 * count / n)
+
+    sc.stop()

examples/src/main/python/sort.py

@@ -34,3 +34,5 @@
     output = sortedCount.collect()
     for (num, unitcount) in output:
         print num
+
+    sc.stop()

examples/src/main/python/transitive_closure.py

@@ -64,3 +64,5 @@ def generateGraph():
             break
 
     print "TC has %i edges" % tc.count()
+
+    sc.stop()

examples/src/main/python/wordcount.py

@@ -33,3 +33,5 @@
     output = counts.collect()
     for (word, count) in output:
         print "%s: %i" % (word, count)
+
+    sc.stop()