For each partition, apply a function to the set of loci assigned to that partition as well as all regions that overlap those loci (possibly with a grace-window baked in upstream; see PartitionedRegions constructors below).
For each partition, apply a function to the set of loci assigned to that partition as well as all regions that overlap those loci (possibly with a grace-window baked in upstream; see PartitionedRegions constructors below).
function that operates on the regions and loci corresponding to a partition, emiting an output iterator of arbitrary type V.
RDD[V], with partitions comprised of the Iterator[V]'s returned by application of f
to each
partition.
Write the partitioned regions RDD to a file.
Groups a LociPartitioning with an RDD[ReferenceRegion] that has already been partitioned according to that partitioning.
This means some regions will occur multiple times in the RDD (due to regions straddling partition boundaries), so it's important not to confuse this with a regular RDD[ReferenceRegion].
The main API exposed here is mapPartitions, which lets the caller apply a function to a LociSet as well as all region copies that overlap those loci.
Note: the containing PartitionedRegions gets picked up by the closure-cleaner and serialized when mapPartitions is called.