NAME

dgord - compute sparse matrix orderings of graphs in parallel

SYNOPSIS

dgord [options] [gfile] [ofile] [lfile]

DESCRIPTION

The dgord program computes, in a parallel way, an ordering of a Scotch source graph representing the pattern of some symmetric sparse matrix.

Source graph file gfile is either a centralized graph file, or a set of files representing fragments of a distributed graph. The resulting ordering is stored in file ofile. Eventual logging information (such as the one produced by option -v) is sent to file lfile. When file names are not specified, data is read from standard input and written to standard output. Standard streams can also be explicitely represented by a dash '-'.

When the proper libraries have been included at compile time, dgord can directly handle compressed graphs, both as input and output. A stream is treated as compressed whenever its name is postfixed with a compressed file extension, such as in 'brol.grf.bz2' or '-.gz'. The compression formats which can be supported are the bzip2 format ('.bz2'), the gzip format ('.gz'), and the lzma format ('.lzma', on input only).

dgord bases on implementations of the MPI interface to spread work across the processing elements. It is therefore not likely to be run directly, but instead through some launcher command such as mpirun.

OPTIONS

-c?opt: Choose default ordering strategy according to one or several options among:

b: enforce load balance as much as possible.
q: privilege quality over speed (default).
s: privilege speed over quality.
t: enforce safety.
x: enforce scalability.

-h: Display some help.
-m?mfile: Save column block mapping data to file mfile. Mapping data specifies, for each vertex, the index of the column block to which this vertex belongs.
-o?strat: Use parallel graph ordering strategy strat (see PT-Scotch user's manual for more information).
-r?pnum: Set root process for centralized files (default is 0).
-t?tfile: Save partitioning tree data to file tfile. Partitioning tree data specifies, for each vertex, the index of the first vertex of the parent block of the block to which the vertex belongs. Altogether with the mapping data provided in file mfile, it allows one to rebuild the separator tree of the nested dissection process.
-V: Display program version and copyright.
-v?verb: Set verbose mode to verb. It is a set of one of more characters which can be:

s: strategy information.
t: timing information.

EXAMPLES

Run dgord on 5 processing elements to reorder matrix graph brol.grf and save the resulting ordering to file brol.ord, using the default sequential graph ordering strategy:

    $ mpirun -np 5 dgord brol.grf brol.ord

Run dgord on 5 processing elements to reorder the distributed matrix stored on graph fragment files brol5-0.dgr to brol5-4.dgr, and save the resulting ordering to file brol.ord (see dgscat(1) for an explanation of the '%p' and '%r' sequences in names of distributed graph fragments).

    $ mpirun -np 5 dgord brol%p-%r.dgr brol.ord

AUTHOR