Helpful Java Spark library.

Pipelines:

• ConcatColumns - When you need to concat a bunch of columns together
• DropColumns - Pipelines do get messy
• ExplodeVector - All those models sure like results in vectors
• FloorCeiling - For putting caps on your variables
• MultiStringIndexer - For the lazy
• TopCategories - When you want to limit the categorical variables
• XGBoostEstimator - Just using the Scala one, but added all the missing "sets" and "gets"

Params:

• MapParam - When you just need to use a Map

Misc:

• FindContinuousColumns - Dataset helper
• FindCategoricalColumns - Dataset helper
• FindNullColumns - Annoyingly, I have some datasets where columns can always be NULL.
• SmartUnion - For when your columns don't line up and the datasets don't always have the same ones.
• JavaToScala helpers
• SparkUnitTest helpers
• SparkSettings sysouts
• ...