You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
225 lines
9.7 KiB
225 lines
9.7 KiB
2 years ago
|
############################################################################## Copyright(C) 1999-2020 National Technology & Engineering Solutions
|
||
|
|
############################################################################## of Sandia, LLC (NTESS). Under the terms of Contract DE-NA0003525 with
|
||
|
|
############################################################################## NTESS, the U.S. Government retains certain rights in this software.
|
||
|
|
##############################################################################
|
||
|
|
############################################################################## See packages/seacas/LICENSE for details
|
||
|
|
|
||
|
|
##############################################################################
|
||
|
|
# Lines beginning with a "#" are considered a comment.
|
||
|
|
#
|
||
|
|
# Blank lines are ignored.
|
||
|
|
#
|
||
|
|
# This file is set up as a number of lines, each of which consists of
|
||
|
|
# a key phrase, followed by a value.
|
||
|
|
#
|
||
|
|
# The order of the key phrases is not significant
|
||
|
|
#
|
||
|
|
# Case is not significant, unless it is significant in the value of a
|
||
|
|
# variable (such as file names).
|
||
|
|
#
|
||
|
|
# Order of the suboptions is not significant.
|
||
|
|
##############################################################################
|
||
|
|
|
||
|
|
#+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
|
||
|
|
# INPUT EXODUSII FILE = <filename>
|
||
|
|
#
|
||
|
|
# This line contains the name of the input ExodusII mesh file which is
|
||
|
|
# to be load balanced
|
||
|
|
#-----------------------------------------------------------------------------
|
||
|
|
INPUT EXODUSII FILE = testa.exoII
|
||
|
|
|
||
|
|
#+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
|
||
|
|
# OUTPUT NEMESISI FILE = <filename>
|
||
|
|
#
|
||
|
|
# This line is OPTIONAL.
|
||
|
|
#
|
||
|
|
# This line contains the name of the output NemsisI. This filename must
|
||
|
|
# be different than the input filename.
|
||
|
|
#
|
||
|
|
# Default: nem_slice generates a name
|
||
|
|
#-----------------------------------------------------------------------------
|
||
|
|
OUTPUT NEMESISI FILE = testa.nemI
|
||
|
|
|
||
|
|
#+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
|
||
|
|
# OUTPUT VISUALIZATION FILE = <option>
|
||
|
|
#
|
||
|
|
# This line is OPTIONAL.
|
||
|
|
#
|
||
|
|
# Specify if a visualization file should be generated for this decomposition.
|
||
|
|
# The options for this line are "yes", "true", "false", and "no".
|
||
|
|
#
|
||
|
|
# Default: no visualization file
|
||
|
|
#-----------------------------------------------------------------------------
|
||
|
|
OUTPUT VISUALIZATION FILE = true
|
||
|
|
|
||
|
|
#+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
|
||
|
|
# GRAPH TYPE FILE = [ELEMENTAL | NODAL]
|
||
|
|
#
|
||
|
|
# This line determines what kind of a decomposition is going to be
|
||
|
|
# generated. The options are ELEMENTAL and NODAL.
|
||
|
|
#-----------------------------------------------------------------------------
|
||
|
|
GRAPH TYPE = ELEMENTAL
|
||
|
|
|
||
|
|
#+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
|
||
|
|
# DECOMPOSITION METHOD = <options>
|
||
|
|
#
|
||
|
|
# This line contains all of the information about the decomposition
|
||
|
|
# method that will be used. The options are:
|
||
|
|
# multikl - use multilevel Kernighan-Lin partitioning
|
||
|
|
# method
|
||
|
|
# spectral - use spectral partitioning method
|
||
|
|
# scattered - use scattered partitioning method
|
||
|
|
# linear - use linear partitioning method
|
||
|
|
# inertial - use inertial partitioning method
|
||
|
|
# random - use random partitioning method
|
||
|
|
# infile=<filename> - read assignment vector from file
|
||
|
|
# NOTE: one and only one of the above options must be specified
|
||
|
|
#
|
||
|
|
# kl - OPTIONAL, use Kernighan-Lin refinement for
|
||
|
|
# multilevel methods; Default = none
|
||
|
|
# none - OPTIONAL, do not use any refinement;
|
||
|
|
# Default = none
|
||
|
|
# num_sects=<integer> - OPTIONAL, number of eigenvectors; Default = 1
|
||
|
|
# cnctd_dom - OPTIONAL, set CONNECTED_DOMAINS parameter
|
||
|
|
# in Chaco. This option forces Chaco to make
|
||
|
|
# sure that all domains are connected after
|
||
|
|
# the partitioning. Default = off
|
||
|
|
# outfile=<filename> - OPTIONAL, One the partition is complete,
|
||
|
|
# write out the assignment vector to an ascii
|
||
|
|
# file. The nemesisI file is still created if
|
||
|
|
# this option is used. NOTE: This option and
|
||
|
|
# the "infile" option cannot be used at the
|
||
|
|
# same time. Default = off
|
||
|
|
#-----------------------------------------------------------------------------
|
||
|
|
DECOMPOSITION METHOD = INERTIAL,KL, NUM_SECTS=1, CNCTD_DOM
|
||
|
|
|
||
|
|
#+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
|
||
|
|
# SOLVER SPECIFICATIONS = <options>
|
||
|
|
#
|
||
|
|
# This line is OPTIONAL.
|
||
|
|
#
|
||
|
|
# This line contains all of the options for the eigensolver. The options are:
|
||
|
|
# tolerance=<number> - OPTIONAL, set the eigensolver tolerance;
|
||
|
|
# Default = 1.0e-3
|
||
|
|
# use_rqi - OPTIONAL, use the RQI/Symmlq eigensolver;
|
||
|
|
# Default = OFF
|
||
|
|
# vmax=<integer> - OPTIONAL, set the number of vertices to
|
||
|
|
# coarsen down to; Default = 200
|
||
|
|
#-----------------------------------------------------------------------------
|
||
|
|
SOLVER SPECIFICATIONS = TOLERANCE=2.0e-3,USE_RQI,VMAX=200
|
||
|
|
|
||
|
|
#+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
|
||
|
|
# MACHINE DESCRIPTION = <options>
|
||
|
|
#
|
||
|
|
# This line is used to describe the machine. The options are:
|
||
|
|
# hcube=<integer> - Set machine type to hypercube where the
|
||
|
|
# order of the cube is given.
|
||
|
|
# hypercube=<integer> - same as "hcube"
|
||
|
|
# mesh=<mesh description> - Set the machine type to a mesh. The
|
||
|
|
# description can be given as a 1, 2, or
|
||
|
|
# three dimensional mesh by giving the
|
||
|
|
# number of processors in each dimension
|
||
|
|
# separated by an "x". For example, a 2d
|
||
|
|
# mesh with 10 processors on each side is
|
||
|
|
# designated be "10x10"
|
||
|
|
# cluster=<cluster desc> - Set the machine type to a Cluster machine.
|
||
|
|
# This causes nem_slice to slice first on
|
||
|
|
# the number of boxes, and then call Chaco
|
||
|
|
# for each box. The hope is that this will
|
||
|
|
# minimize the communication across the boxes.
|
||
|
|
# The machine description is as follows:
|
||
|
|
# <# boxes><m|h><description>
|
||
|
|
# # boxes - number of boxes, integer
|
||
|
|
# m | h - designates each box as
|
||
|
|
# either mesh or hypercube
|
||
|
|
# description - the description of the
|
||
|
|
# mesh or hypercube same
|
||
|
|
# as above.
|
||
|
|
# example:
|
||
|
|
# cluster=16m8x8
|
||
|
|
# 16 boxes, and each box has a 2d mesh
|
||
|
|
# that is 8 by 8 processors.
|
||
|
|
#-----------------------------------------------------------------------------
|
||
|
|
MACHINE DESCRIPTION = MESH=2x2
|
||
|
|
|
||
|
|
#+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
|
||
|
|
# WEIGHTING SPECIFICATIONS = <options>
|
||
|
|
#
|
||
|
|
# This line is OPTIONAL.
|
||
|
|
#
|
||
|
|
# This line contains any weighting information. If this line is absent,
|
||
|
|
# then all weighting is turned off. The options are:
|
||
|
|
# read=<filename> - OPTIONAL, This is the name of an ExodusII
|
||
|
|
# file from which to read the vertex weights.
|
||
|
|
# If this is specified, then either var_name
|
||
|
|
# or var_index must be specified.
|
||
|
|
# var_name=<var name> - This is the name of the variable in the
|
||
|
|
# file, read, which will be used as vertex
|
||
|
|
# weights.
|
||
|
|
# var_index=<integer> - This is the index of the variable in the
|
||
|
|
# file, read, which will be used as vertex
|
||
|
|
# weights.
|
||
|
|
# time_index=<integer> - OPTIONAL, This is the time index of the
|
||
|
|
# variable to be read. If it is not given
|
||
|
|
# then the default is time index 1.
|
||
|
|
# eb=<block id>:<weight> - OPTIONAL, This is a way to weight a specific
|
||
|
|
# element block. Weights should be specified
|
||
|
|
# with integers. So, in order to assign element
|
||
|
|
# block id 10 with a weight of 6 the following
|
||
|
|
# would be used: eb=10:6. If an element block
|
||
|
|
# weight is place after a read on the weighting
|
||
|
|
# specification line, then it will overwrite
|
||
|
|
# the vertex weights from the read.
|
||
|
|
# edges - turn edge weights on
|
||
|
|
#-----------------------------------------------------------------------------
|
||
|
|
WEIGHTING SPECIFICATIONS = READ=testa-out-r.exoII,Var_Name=Temp,time_index=5
|
||
|
|
|
||
|
|
#+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
|
||
|
|
# MISC OPTIONS = <options>
|
||
|
|
#
|
||
|
|
# This line is OPTIONAL.
|
||
|
|
#
|
||
|
|
# This line is used to handle any miscellaneous information that doesn't
|
||
|
|
# really fit anywhere else. The options are:
|
||
|
|
# checks_off - OPTIONAL, This turns off some of the
|
||
|
|
# error checking that nem_slice performs
|
||
|
|
# when it is finding an elements side ids.
|
||
|
|
# NOTE: this option should be used with
|
||
|
|
# caution.
|
||
|
|
# face_adj - OPTIONAL, In an elemental distribution,
|
||
|
|
# use a face concept of adjacency to generate
|
||
|
|
# the graph that will be passed to Chaco.
|
||
|
|
# This definition only allows adjacency
|
||
|
|
# between elements if they share an entire
|
||
|
|
# face. The default is for elements to be
|
||
|
|
# adjacent if the share at least one node.
|
||
|
|
# groups {group designator} - Define groups that will be passed into
|
||
|
|
# Chaco separately. The groups are designated
|
||
|
|
# by element block ids. This option should be
|
||
|
|
# used when there is a problem handling
|
||
|
|
# elements from different blocks on the same
|
||
|
|
# processor. The rules for the group designator
|
||
|
|
# are as follows:
|
||
|
|
# - The beginning and end of the group
|
||
|
|
# designator are set using brackets "{","}"
|
||
|
|
# - Blocks are grouped using the slash "/"
|
||
|
|
# character
|
||
|
|
# - Ids are separated with white space, comma,
|
||
|
|
# or by the hyphen "-" character
|
||
|
|
# - Any blocks not included in the list, are
|
||
|
|
# added to a separate group
|
||
|
|
# - Duplicates in the list are permitted, but
|
||
|
|
# the last group to which a block is placed
|
||
|
|
# is where the block will go
|
||
|
|
# - Block IDs not in the exodus file are
|
||
|
|
# quietly ignored
|
||
|
|
# Example:
|
||
|
|
# Assume block IDs= 1-20 31-45
|
||
|
|
#
|
||
|
|
# descriptor group1 group2 group3
|
||
|
|
# - {1-20} 1-20 31-45
|
||
|
|
# - {30-45 3/ 10-12} 3, 30-45 10,11,12 1,2,4-20
|
||
|
|
# - {1-20/40-45/5-10 21-41} 1-4,11-20 42,43,45 5-10 31-41
|
||
|
|
#-----------------------------------------------------------------------------
|
||
|
|
MISC OPTIONS = checks_off, face_adj, groups {1 / 2}
|