To perform k-means clustering and parallelizing it through MPI.
Value of k is asked from user.
iris.data is given as a dataset. Read the data from file and ignore the last column of each row in dataset.
To compile:
mpicc -o k_means k_means.c
where k_means is the name of the object file.
To run:
mpiexec -n 4 ./k_means
where 4 is the number of processors. You can give any positive integer value to n. But generally machines have 4 processors, so giving value of n greater 4 will create lot of overhead and actually increase the time taken.