[SPARK-13457] Remove DataFrame RDD operations

• Type: Improvement
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 2.0.0
• Fix Version/s: 2.0.0
• Component/s: SQL
• Target Version/s: 2.0.0
• Assignee: Cheng Lian
• Reporter: Cheng Lian
• Created: 23/Feb/16 16:15
• Updated: 26/Feb/16 16:29
• Resolved: 26/Feb/16 16:29

Description

We'd like to remove DataFrame RDD operations like map, filter, and foreach because:

1. After making DataFrame a subclass of Dataset[Row], these methods conflicts with methods in Dataset.
2. By returning RDDs, they are semantically improper.

It's trivial to remove them since they simply delegates to methods of DataFrame.rdd.

Issue Links

• blocks: SPARK-13244 Unify DataFrame and Dataset API
• links to: [Github] Pull Request #11323 (liancheng)
• links to: [Github] Pull Request #11388 (liancheng)