Can I append a new transformer or estimator to the pipeline?

Question

In Spark ML, Can I load saved pipeline and append a new transformer to this loaded pipeline?

Answer 1

You can use the getStages method on the Pipeline object to get an Array[PipelineStage] then append the desired PipelineStage using the :+ scala operator.

val pipeline = //your pipeline
val newStage = //you new stage
pipeline.getStages :+ new Stage

Hope it helps. Best Regards