Parallel-K-Means-Clustering

Sequential and Parallel(using Open MP and Pthreads) Implementations(c++) of the K Means Clustering Algorithm and visualizing the results for a comparative study of the Speedup and Efficiency achieved in 3 different implementations

Execution

• Sequential

  • cd into Sequential folder
  • g++ main_sequential.c lab1_io.c Kmeans-Sequential.cpp -fopenmp -o seq.out
  • ./seq.out 4 sample_dataset_50000_4.txt b.txt c.txt
  • 4 is for the number of clusters. Can be changed.
  • To Visualize the results :- python visualise.py b.txt

• OpenMP

  • cd into OpenMP folder
  • g++ main_omp.c lab1_io.c Kmeans-OpenMP.cpp -fopenmp -o omp.out
  • ./omp.out 4 2 sample_dataset_5000_3.txt b.txt c.txt
  • 4 is for the number of clusters and 2 is for the number of threads. Both can be changed.
  • To Visualize the results :- python visualise.py b.txt

• P-Threads

  • cd into P-Threads folder
  • g++ lab1_io.c main_pthread.c Kmeans-Pthreads.cpp -o ptry.out -fopenmp
  • ./ptry.out 4 2 sample_dataset_50000_4.txt b.txt c.txt
  • 4 is for the number of clusters and 2 is for the number of threads. Both can be changed.
  • To Visualize the results :- python visualise.py b.txt

Sample Visualization