A map/reduce-based computing framework implemented by Python using multiprocessing. The first version only facilicates from the multi-cores of a single machine. New version will take multiple machines into accound by delivering messages among machines across the network.

The procedure of a map/reduce job is as follows:

masive input data -> splitter -> data partitions -> mapper(sorter) -> shuffler -> reducer -> collector -> output data