run_droplet_test module

This example uses three modules instantiated many times in two different networks. Each network configuration uses a different amount of module instances and a different network topology. This example can be executed with MPI (if mpi4py is available) or with the Python multiprocessing library. These choices are made by variables listed below in the executable portion of this run file.

Single Plot

The first network case is named “single plot”. Here one DataPlot module is connected to all Droplet modules. To run this case using MPI you should compute the number of processes as follows:

nprocs = n_droplets + 1 vortex + 1 data_plot + 1 cortix

then issue the MPI run command as follows (replace nprocs with a number):

mpiexec -n nprocs run_droplet.py

To run this case with the Python multiprocessing library, just run this file at the command line as

run_droplet.py

Multiple Plot

The second network case is named “multiple plot”. Here each Droplet is connected to an instance of the DataPlot module, therefore many more nodes are added to the network when compared to the first network case. To run this case using MPI compute

nprocs = 2*n_droplets + 1 vortex + 1 cortix

then issue the MPI run command as follows (replace nprocs:

mpiexec -n nprocs run_droplet.py

To run this case with the Python multiprocessing library, just run this file at the command line as

run_droplet.py