Type Members

1.  trait CrossSparkUtils extends AnyRef
2.  type Dataset[X] = sql.Dataset[X]
3.  class DefaultSource extends RelationProvider with SchemaRelationProvider with CreatableRelationProvider with DataSourceRegister

   Provides access to H2OFrame from pure SQL statements (i.e.

   Provides access to H2OFrame from pure SQL statements (i.e. for users of the JDBC server).

4.  type Frame = water.fvec.Frame
5.  type H2O = water.H2O
6.  class H2OConf extends Logging with InternalBackendConf with ExternalBackendConf with Serializable

   Configuration holder which is representing properties passed from user to Sparkling Water.

7.  class H2OContext extends H2OContextExtensions

   Create new H2OContext based on provided H2O configuration

8.  implicit class H2ODataFrameReader extends AnyRef

   Adds a method, h2o, to DataFrameReader that allows you to read h2o frames using the DataFileReader.

   Adds a method, h2o, to DataFrameReader that allows you to read h2o frames using the DataFileReader. It's alias for sqlContext.read.format("org.apache.spark.h2o").option("key",frame.key.toString).load()

9.  implicit class H2ODataFrameWriter[T] extends AnyRef

   Adds a method, h2o, to DataFrameWriter that allows you to write h2o frames using the DataFileWriter.

   Adds a method, h2o, to DataFrameWriter that allows you to write h2o frames using the DataFileWriter. It's alias for sqlContext.write.format("org.apache.spark.h2o").option("key","new_frame_key").save()

10.  type H2OFrame = water.fvec.H2OFrame
11.  class JavaH2OContext extends AnyRef

    A Java-friendly version of org.apache.spark.h2o.H2OContext

    A Java-friendly version of org.apache.spark.h2o.H2OContext

    Sparkling Water can run in two modes. External cluster mode and internal cluster mode. When using external cluster mode, it tries to connect to existing H2O cluster using the provided spark configuration properties. In the case of internal cluster mode,it creates H2O cluster living in Spark - that means that each Spark executor will have one h2o instance running in it. This mode is not recommended for big clusters and clusters where Spark executors are not stable.

    Cluster mode can be set using the spark configuration property spark.ext.h2o.mode which can be set in script starting sparkling-water or can be set in H2O configuration class H2OConf

12.  type RDD[X] = rdd.RDD[X]
13.  class WrongSparkVersion extends Exception with NoStackTrace