[SPARK-13340] [ML] PolynomialExpansion and Normalizer should validate input type

• Type: Bug
• Status: Resolved
• Priority: Trivial
• Resolution: Incomplete
• Affects Version/s: 1.6.0, 2.0.0
• Component/s: ML, MLlib
• Labels: bulk-closed
• Assignee: Grzegorz Chilkiewicz
• Reporter: Grzegorz Chilkiewicz
• Shepherd: Nicholas Pentreath
• Created: 16/Feb/16 13:50
• Updated: 25/May/21 01:54
• Resolved: 25/May/21 01:41

Description

PolynomialExpansion and Normalizer should override UnaryTransformer::validateInputType

Now, in case of trying to operate on String column:
java.lang.ClassCastException: java.lang.String cannot be cast to org.apache.spark.mllib.linalg.Vector
is thrown, but:
java.lang.IllegalArgumentException: requirement failed: Input type must be VectorUDT but got StringType
will be more clear and adequate

Issue Links

• links to: [Github] Pull Request #11218 (grzegorz-chilkiewicz)
• links to: GitHub Pull Request #11218