[SPARK-2514] [mllib] Random RDD generator

[dorx]

Utilities for generating random RDDs.

RandomRDD and RandomVectorRDD are created instead of using sc.parallelize(range:Range) because Range objects in Scala can only have size <= Int.MaxValue.

The object RandomRDDGenerators can be transformed into a generator class to reduce the number of auxiliary methods for optional arguments.

[SparkQA]

QA tests have started for PR 1520. This patch merges cleanly.
View progress: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/16942/consoleFull

[dorx]

@falaki @jkbradley @mengxr

[SparkQA]

QA results for PR 1520:
- This patch PASSES unit tests.
- This patch merges cleanly
- This patch adds the following public classes (experimental):
trait DistributionGenerator extends Pseudorandom with Serializable {
class UniformGenerator() extends DistributionGenerator {
class StandardNormalGenerator() extends DistributionGenerator {
class PoissonGenerator(val mean: Double) extends DistributionGenerator {

For more information see test ouptut:
https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/16942/consoleFull

[mengxr]

Change @return to Returns. Otherwise the summary will be empty in the generated docs.

[SparkQA]

QA tests have started for PR 1520. This patch merges cleanly.
View progress: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/17060/consoleFull

[mengxr]

i.i.d -> i.i.d. and in other places

[mengxr]

@dorx Besides comments, could you mark distribution generators and methods that requires distribution generators @Experimental? Part of the reason is that we don't have the API in Python and whether we should implement the same in Python is not clear.

[SparkQA]

QA tests have started for PR 1520. This patch merges cleanly.
View progress: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/17197/consoleFull

[SparkQA]

QA results for PR 1520:
- This patch FAILED unit tests.
- This patch merges cleanly
- This patch adds the following public classes (experimental):
trait DistributionGenerator extends Pseudorandom with Serializable {
class UniformGenerator extends DistributionGenerator {
class StandardNormalGenerator extends DistributionGenerator {
class PoissonGenerator(val mean: Double) extends DistributionGenerator {

For more information see test ouptut:
https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/17197/consoleFull

[dorx]

Jenkins, retest this please.

[SparkQA]

QA tests have started for PR 1520. This patch merges cleanly.
View progress: https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/17205/consoleFull

[SparkQA]

QA results for PR 1520:
- This patch PASSES unit tests.
- This patch merges cleanly
- This patch adds the following public classes (experimental):
trait DistributionGenerator extends Pseudorandom with Serializable {
class UniformGenerator extends DistributionGenerator {
class StandardNormalGenerator extends DistributionGenerator {
class PoissonGenerator(val mean: Double) extends DistributionGenerator {

For more information see test ouptut:
https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/17205/consoleFull

[mengxr]

LGTM. Merged into master. Thanks for adding random RDD generators!!