Changeset 2152 for inundation/parallel/run_parallel_sw_rectangle.py

Timestamp:

Dec 16, 2005, 9:59:58 AM (19 years ago)

Author:

linda

Message:

Added caching to the sw_merimbula files. Also updated some of the comments

File:

• inundation/parallel/run_parallel_sw_rectangle.py (modified) (4 diffs)

inundation/parallel/run_parallel_sw_rectangle.py

@@ -1 @@
-#!/usr/bin/env/python
+#!/usr/bin/env python
 #########################################################
 #
@@ -7 +7 @@
 # file.
 #
-#  *) The test files currently avaliable are of the form
-# test*.out, eg test_5l_4c.out. The term infront of the l
-# corresponds to the number of levels of refinement
-# required to build the grid, i.e. a higher number
-# corresponds to a finer grid. The term infront of the c
-# corresponds to the number of processors.
 #
-# *) The (new) files that have been added to manage the
-# grid partitioning are
-#    +) pmesh_divide.py: subdivide a pmesh
-#    +) build_submesh.py: build the submeshes on the host
-# processor.
-#    +) build_local.py: build the GA mesh datastructure
-# on each processor.
-#    +) build_commun.py: handle the communication between
-# the host and processors
-#
-# *) Things still to do:
-#    +) Overlap the communication and computation: The
-# communication routines in build_commun.py should be
-# interdispersed in the build_submesh.py and build_local.py
-# files. This will overlap the communication and
-# computation and will be far more efficient. This should
-# be done after more testing and there more confidence in
-# the subpartioning.
-#    +) Much more testing especially with large numbers of
-# processors.
 #  Authors: Linda Stals, Steve Roberts and Matthew Hardy,
 # June 2005
@@ -62 +36 @@
 
 from pmesh_divide import pmesh_divide, pmesh_divide_steve
-from build_submesh import *
-from build_local import *
-from build_commun import *
-from pmesh2domain import pmesh_to_domain_instance
 
 # read in the processor information
@@ -72 +42 @@
 myid = pypar.rank()
 processor_name = pypar.Get_processor_name()
-
-
-
-
 
 M = 20