Package org.apache.sysds.runtime.instructions.spark.data

Class Summary

Class	Description
BlockPartitioner	Default partitioner used for all binary block rdd operations in order to enable sufficient local aggregation independent of the aggregation direction (row/col-wise).
BroadcastObject<T extends CacheBlock>
CorrMatrixBlock
DatasetObject
FrameReblockBuffer
IndexedMatrixValue
LazyIterableIterator<T>	This class is a generic base class for lazy, single pass iterator classes in order to simplify the implementation of lazy iterators for mapPartitions use cases.
LineageObject
PartitionedBlock<T extends CacheBlock>	This class is for partitioned matrix/frame blocks, to be used as broadcasts.
PartitionedBroadcast<T extends CacheBlock>	This class is a wrapper around an array of broadcasts of partitioned matrix/frame blocks, which is required due to 2GB limitations of Spark's broadcast handling.
RDDObject
ReblockBuffer
RowMatrixBlock
SerLongWritable	Wrapper for LongWritable in order to make it serializable as required for shuffle in spark instructions.
SerText	Wrapper for Text in order to make it serializable as required for shuffle in spark instructions.