chi.LinearGraphPartitioner

Basic linear partitioning. This type of partitioner works basically only for testing. Orthogonal meshes can produce decent partitioning but for unstructured grids it can be pretty bad. It partitions cells based on their linear index "global_id" instead of actually working with the graph.

Example usage:

Create this object:

params = 
{
    param_name1 = value1,
    param_name2 = value2,
    --etc.
}
chi.LinearGraphPartitioner.Create(params)

Optional Input parameters

►all_to_rank       If non-zero will restrict all cells to this rank, essentially transforming this partitioner into a single-rank partitioner

type=INTEGER. If non-zero will restrict all cells to this rank, essentially transforming this partitioner into a single-rank partitioner.

Default value: -1