[FLINK-2984] Support lenient parsing of SVMLight input files

• Type: Improvement
• Status: Closed
• Priority: Minor
• Resolution: Implemented
• Affects Version/s: 0.9.1
• Fix Version/s: 1.0.0
• Component/s: Library / Machine Learning
• Labels: easyfix
• Assignee: Chiwan Park
• Reporter: Johannes
• Created: 06/Nov/15 13:48
• Updated: 13/Apr/21 20:39
• Resolved: 15/Jan/16 01:22

Description

The current implementation for the reader assumes that the format follows the exact specification.

The splice-site Dataset dataset is formatted slightly different

Example

-1  1:0.381846 2:0.163648 3:0.245472 4:0.627318

note the two spaces after the label.

Currently MLUtils.scala splits on single spaces.

Issue Links

• is part of: FLINK-1717 Add support to read libSVM and SVMLight input files