MPI_INT - Integer
MPI_DOUBLE - 8 byte floating point number
We have support for scalars and multiple dimension arrays using the Numeric package. The package can also be built using NumPy instead of Numeric. See the README file.