When parallel tasks are independent from each other, they are called embarrassingly parallel. While there are several ways to approach these problems, including batch schedulers, we can also write simple MPI programs to complete these tasks.
In Homework Assignment 8 we wrote a program that would process a data file, converting the raw data to true stress and strain, and compute the toughness. If we had multiple data files, we could use this program to process them in parallel on a multi-processor machine. In the assignment that follows, we will only do this for two data files, but of course, the real utility of writing a parallel program would only be realized if we had many more to process.
Your assignment is to complete the Julia module assignment11.jl. Specifically, you must complete the function compute_toughness_parallel(path_to_data::String, comm::MPI.Comm) where path_to_data is a string that contains the file directory path (not the path to any particular file, just the directory) and comm is an MPI.COMM_WORLD object. When the script is executed with the following command:
mpiexecjl --project=. -np 2 julia src/assignment11.jl datafrom the assignment root directory, the script should process the two data files data_0.csv and data_1.csv in parallel. I am including my solution to Homework Assignment 8 as part of the existing assignment11.jl. The data argument in the command above is the path to the data files.
Please run the following commands from the Julia REPL one time before starting your assignment to install the project aware version of mpiexec.
using MPI; MPI.install_mpiexecjl()To see if you answers are correct, run the following command at the Terminal command line from the repository's root directory
julia --project=. -e "using Pkg; Pkg.test()"the tests will run and report if passing or failing.