[SPARK-5893] Add Bucketizer

• Type: Sub-task
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 1.4.0
• Component/s: ML
• Target Version/s: 1.4.0
• Assignee: Xusen Yin
• Reporter: Xiangrui Meng
• Created: 18/Feb/15 19:28
• Updated: 13/May/15 01:08
• Resolved: 12/May/15 01:41

Description

`Bucketizer` maps a column of continuous features to a column of feature buckets, where the buckets are specified by users.

val bucketizer = new Bucketizer()
  .setInputCol("age")
  .setBuckets(Array(10, 20, 30, 40, 50))
  .setOutputCol("ageBuckets")

The output feature names should be automatically generated. We need to discuss the behavior if a value falls out the specified bucket range.

Issue Links

• is depended upon by: SPARK-7593 Python API for Bucketizer
• relates to: SPARK-5890 Add QuantileDiscretizer
• relates to: SPARK-7577 User guide update for Bucketizer
• links to: [Github] Pull Request #5980 (yinxusen)