[SPARK-16965] Fix bound checking for SparseVector

• Type: Bug
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 2.0.0
• Fix Version/s: 2.1.0
• Component/s: MLlib, PySpark
• Target Version/s: 2.1.0
• Assignee: Jeff Zhang
• Reporter: Jeff Zhang
• Created: 09/Aug/16 05:29
• Updated: 19/Aug/16 11:38
• Resolved: 19/Aug/16 11:38

Description

There's several issues in the bound checking of SparseVector

1. In scala, miss negative index checking and different bound checking is scattered in several places. Should put them in one place
2. In python, miss low/upper bound checking of indices.

Issue Links

• is duplicated by: SPARK-17130 SparseVectors.apply and SparseVectors.toArray have different returns when creating with a illegal indices
• links to: [Github] Pull Request #14555 (zjffdu)