[SPARK-14873] Java sampleByKey methods take ju.Map but with Scala Double values; results in type Object

• Type: Sub-task
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 1.6.1
• Fix Version/s: 2.0.0
• Component/s: Java API, Spark Core
• Target Version/s: 2.0.0
• Assignee: Sean R. Owen
• Reporter: Sean R. Owen
• Created: 23/Apr/16 09:23
• Updated: 23/Apr/16 17:47
• Resolved: 23/Apr/16 17:47

Description

There's this odd bit of code in JavaStratifiedSamplingExample:

    // specify the exact fraction desired from each key Map<K, Object>
    ImmutableMap<Integer, Object> fractions =
      ImmutableMap.of(1, (Object)0.1, 2, (Object) 0.6, 3, (Object) 0.3);

    // Get an approximate sample from each stratum
    JavaPairRDD<Integer, Character> approxSample = data.sampleByKey(false, fractions);

It highlights a problem like that in https://issues.apache.org/jira/browse/SPARK-12604 where Scala primitive types are used where Java requires an object, and the result is that a signature that logically takes Double (objects) takes an Object in the Java API. It's an easy, similar fix.

Issue Links

• duplicates: SPARK-12604 Java count(AprroxDistinct)ByKey methods return Scala Long not Java
• links to: [Github] Pull Request #12637 (srowen)