[SPARK-15009] PySpark CountVectorizerModel should be able to construct from vocabulary list

• Type: Improvement
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Fix Version/s: 2.4.0
• Component/s: ML, PySpark
• Target Version/s: 2.4.0
• Assignee: Bryan Cutler
• Reporter: Bryan Cutler
• Created: 29/Apr/16 18:42
• Updated: 29/Mar/18 23:39
• Resolved: 16/Mar/18 18:44

Description

Like the Scala version, PySpark CountVectorizerModel should be able to construct the model from given a vocabulary list.

For example

cvm = CountVectorizerModel(["a", "b", "c"])

Issue Links

• is related to: SPARK-14087 PySpark ML JavaModel does not properly own params after being fit
• is required by: SPARK-15623 2.0 python coverage ml.feature
• is required by: SPARK-16486 Python API parity issues from 2.0 QA
• relates to: SPARK-21812 PySpark ML Models should not depend transfering params from Java
• relates to: SPARK-17161 Add PySpark-ML JavaWrapper convenience function to create py4j JavaArrays
• links to: [Github] Pull Request #16770 (BryanCutler)
• links to: [Github] Pull Request #20942 (BryanCutler)