Skip to content

Latest commit

 

History

History
27 lines (23 loc) · 1.4 KB

README.md

File metadata and controls

27 lines (23 loc) · 1.4 KB
Raw

KNN_Classification

KNN Classification with MPI and openMP support

The source code of the program is in the Source folder. There are 3 versions:

  • The MPI Blocking version MPI_B_KNN.c
  • The MPI Non-Blocking version MPI_NB_KNN.c
  • A serial version Serial_KNN.c

The data used to produce the resulting graphs found on the images folder can be found in the Files directory here:

To compile and run the programs:

  1. download the Source and the Files directories from the Dropbox link and put them on the same directory.
  2. Go to the Source directory (cd path/Source)
  3. Compile the programs:
    • Blocking: mpicc -O3 -fopenmp functions.c MPI_B_KNN.c -o BL
    • Non Blocking: mpicc -O3 -fopenmp functions.c MPI_NB_KNN.c -o NBL
    • Serial: `gcc -O3 Serial_KNN.c -o SE -lrt
  4. Run the programs on a single computer (example with the data provided in the Files directory and 4 procs):
    • Blocking: mpiexec -np 4 ./BL 10000 784
    • Non Blocking: mpiexec -np 4 ./NBL 60000 30

For running the scripts on the hellasgrid cluster you can use scripts like the ones provided in the Scripts folder. The matlab script was used to extract the plots shown at the images folder.