I am a new user of MPI and I would like to have your opinion on the following issue.

With 1 processor I keep the data in

std::vector<std::vector<double> > X

Then I call a function that takes various arguments, among which X[ i ] and X[ j ], i.e. two sub-vectors of X.

With MPI I would have the same function execute on different processors for different values of i and j in X[ i ] and X[ j ].

From my understanding since std::vector is not an MPI type it cannot be passed neither as a message nor an element of an MPI structure. I thought of using

double X[200][4000]

which MPI would pass but I don't see how to isolate a single row, i.e. X[99] would not make sense in this context.

Any suggestion for an efficient and possibly simple solution?

Thank you for your help,

Alessandro