[SPARK-14565] RandomForest should use parseInt and parseDouble for feature subset size instead of regexes

• Type: Improvement
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Fix Version/s: 2.0.0
• Component/s: ML
• Target Version/s: 2.0.0
• Assignee: Yong Tang
• Reporter: Xiangrui Meng
• Shepherd: Xiangrui Meng
• Created: 12/Apr/16 18:08
• Updated: 15/Apr/16 00:23
• Resolved: 15/Apr/16 00:23

Description

Using regex is not robust and hard to maintain.

Issue Links

• relates to: SPARK-3724 RandomForest: More options for feature subset size
• links to: [Github] Pull Request #12360 (yongtang)