Changeset 15


Ignore:
Timestamp:
Nov 4, 2008 4:44:48 PM (11 years ago)
Author:
pbelotti
Message:

some minor updates

Location:
trunk
Files:
2 edited

Legend:

Unmodified
Added
Removed
  • trunk/Couenne/src/couenne.opt

    r14 r15  
    1 display_stats
    2 display statistics at the end of the run
     1#
     2# Couenne options
     3#
     4# Couenne is an open-source solver for nonconvex Mixed-Integer
     5# Nonlinear Programming (MINLP) problems. See more at
     6#
     7# https://projects.coin-or.org/Couenne
     8#
     9# The following is a list of option to tweak the performance of Couenne.
     10# Each option has a brief description and is set to a default value that we believe
     11# works well in most cases. in square brackets are given the possible values
    312
    4 branching_print_level
    5 Output level for braching code in Couenne
    613
    7 boundtightening_print_level
    8 Output level for bound tightening code in Couenne
     14#
     15# Verbosity/output level options
     16#
    917
    10 convexifying_print_level
    11 Output level for convexifying code in Couenne
     18# display statistics at the end of the run [yes/no]
     19display_stats no
    1220
    13 problem_print_level
    14 Output level for problem manipulation code in Couenne
     21# The following are verbosity levels for the main components of Couenne.
     22# Values: 0 for quiet, 11 for excessive output [0..11]
    1523
    16 nlpheur_print_level
    17 Output level for NLP heuristic in Couenne
     24branching_print_level       0 # Output level for braching code in Couenne.
     25boundtightening_print_level 0 # Output level for bound tightening code in Couenne
     26convexifying_print_level    0 # Output level for convexifying code in Couenne
     27problem_print_level         4 # Output level for problem manipulation code in Couenne (4 prints out original problem)
     28nlpheur_print_level         0 # Output level for NLP heuristic in Couenne
     29#disjcuts_print_level       0 # Output level for disjunctive cuts in Couenne - disabled for now
    1830
    19 disjcuts_print_level
    20 Output level for disjunctive cuts in Couenne
    2131
     32#
     33# Option for branching rules
     34#
     35
     36# Multipliers of pseudocosts for estimating and update estimation of bound
    2237pseudocost_mult
    23 Multipliers of pseudocosts for estimating and update estimation of bound
    2438
     39# Use distance between LP points to update multipliers of pseudocosts " 
    2540pseudocost_mult_lp
    26 Use distance between LP points to update multipliers of pseudocosts " 
    2741
     42# Use Special Ordered Sets (SOS) as indicated in the MINLP model
    2843enable_sos
    29 Use Special Ordered Sets (SOS) as indicated in the MINLP model
    3044
     45# Apply bound tightening before branching
    3146branch_fbbt
    32 Apply bound tightening before branching
    3347
     48# Apply convexification cuts before branching (for now only within strong branching)
    3449branch_conv_cuts
    35 Apply convexification cuts before branching (for now only within strong branching)
    3650
     51# Chooses branching point selection strategy
    3752branch_pt_select
    38 Chooses branching point selection strategy
    3953
     54# Defines convex combination of mid point and current LP point: "
    4055branch_midpoint_alpha
    41 Defines convex combination of mid point and current LP point: "
    4256
     57# Defines safe interval percentage for using LP point as a branching point
    4358branch_lp_clamp
    44 Defines safe interval percentage for using LP point as a branching point
    4559
     60# Priority of continuous variable branching
    4661cont_var_priority
    47 Priority of continuous variable branching
    4862
     63# Apply Reduced Cost Branching (instead of the Violation Transfer) -- MUST have vt_obj enabled
    4964red_cost_branching
    50 Apply Reduced Cost Branching (instead of the Violation Transfer) -- MUST have vt_obj enabled
    5165
     66# type of branching object for variable selection
     67branching_object
     68
     69
     70#
     71# Options for bound tightening
     72#
     73
     74# Feasibility-based (cheap) bound tightening
     75feasibility_bt
     76
     77# Optimality-based (expensive) bound tightening
     78optimality_bt
     79
     80log_num_obbt_per_level
     81# Specify the frequency (in terms of nodes) for optimality-based bound tightening.
     82
     83aggressive_fbbt
     84# Aggressive feasibility-based bound tightening (to use with NLP points)
     85
     86log_num_abt_per_level
     87# Specify the frequency (in terms of nodes) for aggressive bound tightening.
     88
     89
     90#
     91# Options for reformulation and linearization
     92#
     93
     94# Specify the frequency (in terms of nodes) at which couenne ecp cuts are generated.
    5295convexification_cuts
    53 Specify the frequency (in terms of nodes) at which couenne ecp cuts are generated.
    5496
    55 check_lp
    56 Check all LPs through an independent call to OsiClpSolverInterface::initialSolve()
    57 
    58 local_optimization_heuristic
    59 Do we search for local solutions of NLP's
    60 
    61 log_num_local_optimization_per_level
    62 Specify the logarithm of the number of local optimizations to perform"
    63 
     97# Determines in which point the linear over/under-estimator are generated
    6498convexification_type
    65 Deterimnes in which point the linear over/under-estimator are generated
    6699
    67100convexification_points
    68 Specify the number of points at which to convexify when convexification type"
     101# Specify the number of points at which to convexify when convexification type"
    69102
    70103violated_cuts_only
    71 Yes if only violated convexification cuts should be added
     104# Yes if only violated convexification cuts should be added
    72105
     106
     107
     108#
     109# Options for debugging
     110#
     111# Some of these options usually slow down Couenne, and are hence only suggested for debugging purposes.
     112#
     113
     114# Check all LPs through an independent call to OsiClpSolverInterface::initialSolve()
     115check_lp
     116
     117# Artificial cutoff
    73118art_cutoff
    74 Artificial cutoff
    75119
     120# Window around known optimum
    76121opt_window
    77 Window around known optimum
    78122
     123# Artificial lower bound
     124art_lower
     125
     126
     127#
     128# Other options
     129#
     130
     131# Do we search for local solutions of NLP's?
     132local_optimization_heuristic
     133
     134# Specify the logarithm of the number of local optimizations to perform
     135log_num_local_optimization_per_level
     136
     137# Tolerance for constraints/auxiliary variables
    79138feas_tolerance
    80 Tolerance for constraints/auxiliary variables
    81139
    82 feasibility_bt
    83 Feasibility-based (cheap) bound tightening
     140# Use quadratic expressions and related exprQuad class
     141use_quadratic
    84142
    85 use_quadratic
    86 Use quadratic expressions and related exprQuad class
    87 
    88 optimality_bt
    89 Optimality-based (expensive) bound tightening
    90 
    91 log_num_obbt_per_level
    92 Specify the frequency (in terms of nodes) for optimality-based bound tightening.
    93 
    94 aggressive_fbbt
    95 Aggressive feasibility-based bound tightening (to use with NLP points)
    96 
    97 log_num_abt_per_level
    98 Specify the frequency (in terms of nodes) for aggressive bound tightening.
    99 
    100 art_lower
    101 Artificial lower bound
    102 
    103 branching_object
    104 type of branching object for variable selection
  • trunk/INSTALL

    r14 r15  
    127127the "g" after the "write" keyword).
    128128
    129 Couenne can be fine tuned by setting parameters in the option file couenne.opt,
     129Couenne can be fine-tuned by setting parameters in the option file couenne.opt,
    130130which is read from the same directory where Couenne is launched. A typical
    131131couenne.opt is provided in Couenne/src/, with some explanation on how to use
    132 its parameters.
     132its parameters. You may copy it to your directory and change some of the parameters,
     133however those set by default are usually best on most of the instances we have tried.
    133134
    134135
Note: See TracChangeset for help on using the changeset viewer.