An Efficient Parallel Triangulation Algorithm for Spherical and Planar Grids with Commonality and Parallel Consistency
- Support for spherical grids and planar grids.
- Compatible with earth system model grids.
- Dynamic expansion with consistency verification.
- MPI and OpenMP parallelism.
- Low computing redundancy.
- High efficiency.
For a quick start:
Just execute make in this directory.
For advance usages:
Some environment variables can be useful, e.g. PAT_OPENCV, PAT_NETCDF, PAT_TIMING and PAT_DEBUG.
The executing command is likely OMP_NUM_THREADS=nt mpiexec -n np ./patcc gridFile.
nt: number of openMP threads.
np: number of MPI processes.
gridFile: a file containing formatted grid info.
At end of the execution, the program will write results to log/global_triangles_* file.
1st line: N, the number of points to read
2nd line: Boundary of the points (minLon maxLon minLat maxLat)
3rd~N+2th lines: coordnate values in degree for each point (lon lat)
A example file named test.dat can be found in this directory.
This project runs as part of C-Coupler