[SPARK-13355] Replace GraphImpl.fromExistingRDDs by Graph

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.3.1, 1.4.1, 1.5.2, 1.6.0, 2.0.0
• Fix Version/s: 1.4.2, 1.5.3, 1.6.1, 2.0.0
• Component/s: ML, MLlib
• Target Version/s: 1.5.3, 1.6.1, 2.0.0
• Assignee: Xiangrui Meng
• Reporter: Xiangrui Meng
• Created: 17/Feb/16 01:46
• Updated: 03/Nov/16 23:48
• Resolved: 23/Feb/16 07:55

Description

`GraphImpl.fromExistingRDDs` expects preprocessed vertex RDD as input. We call it in LDA without validating this requirement. So it might introduce errors. Replacing it by `Gpaph.apply` would be safer and more proper because it is a public API.

Issue Links

• contains: SPARK-12488 LDA describeTopics() Generates Invalid Term IDs
• is related to: SPARK-12363 PowerIterationClustering test case failed if we deprecated KMeans.setRuns
• is required by: SPARK-5572 LDA improvement listing
• links to: [Github] Pull Request #11226 (mengxr)