[SPARK-10959] PySpark StreamingLogisticRegressionWithSGD does not train with given regParam and convergenceTol parameters

• Type: Bug
• Status: Resolved
• Priority: Critical
• Resolution: Fixed
• Affects Version/s: 1.5.1
• Fix Version/s: 1.5.2, 1.6.0
• Component/s: PySpark
• Labels: backport-needed
• Target Version/s: 1.5.2, 1.6.0
• Assignee: Bryan Cutler
• Reporter: Bryan Cutler
• Shepherd: Xiangrui Meng
• Created: 06/Oct/15 21:06
• Updated: 13/Oct/15 22:59
• Resolved: 13/Oct/15 22:59

Description

These parameters are passed into the StreamingLogisticRegressionWithSGD constructor, but do not get transferred to the model to use when training. Same problem with StreamingLinearRegressionWithSGD and the intercept param is in the wrong argument place where it is being used as regularization value.

Issue Links

• links to: [Github] Pull Request #9002 (BryanCutler)
• links to: [Github] Pull Request #9087 (BryanCutler)