Changes between Version 56 and Version 57 of WikiStart


Ignore:
Timestamp:
Apr 10, 2013 2:23:17 PM (5 years ago)
Author:
stefan
Comment:

Cbc 2.8.0 release notes; john still main developer of cbc

Legend:

Unmodified
Added
Removed
Modified
  • WikiStart

    v56 v57  
    1313For more information on supported platforms, links to dependent projects, current version, and more, click [http://www.coin-or.org/projects/Cbc.xml here]
    1414
    15 Cbc was developed by [http://www.fastercoin.com John Forrest], now retired from IBM Research. The project is currently managed by [http://coral.ie.lehigh.edu/~ted/ Ted Ralphs] and the rest of the [CbcTeam Cbc team].
     15Cbc is developed by [http://www.fastercoin.com John Forrest], now retired from IBM Research. The project is currently managed by John Forrest, [http://coral.ie.lehigh.edu/~ted/ Ted Ralphs], and the rest of the [CbcTeam Cbc team].
    1616
    1717-------------
     
    2222 * The latest release version of Cbc is [[Image(htdocs:Cbc-latest-release.png)]].
    2323 * Click [https://projects.coin-or.org/Cbc/log/ here] to see the current change log.
     24
     25April 10, 2013
     26
     27 * Cbc [https://projects.coin-or.org/Cbc/browser/releases/2.8.0 2.8.0] has been released.
     28   * Introduced **new secondaryStatus** 8 to indicate that solving stopped due to an iteration limit.
     29   * **Solution pool** is now accessible via the command line and the !CbcMain* interface.
     30   * New **mipstart option** to read an initial feasible solution from a file.
     31     Only values for discrete variables need to be provided.
     32   * Added **Proximity Search heuristic** by Fischetti and Monaci (off by default):[[BR]]
     33     The simplest way to switch it on using stand-alone version is "-proximity on".[[BR]]
     34     Proximity Search is the new "No-Neighborhood Search" 0-1 MIP refinement heuristic recently proposed by
     35     Fischetti and Monaci (2012). The idea is to define a sub-MIP without additional constraints but with a
     36     modified objective function intended to attract the search in the proximity of the incumbent. The approach
     37     works well for 0-1 MIPs whose solution landscape is not too irregular (meaning the there is reasonable
     38     probability of finding an improved solution by flipping a small number of binary variables), in particular
     39     when it is applied to the first heuristic solutions found at the root node.
     40   * An implementation of **Zero-Half-Cuts** by Alberto Caprara is now available.[[BR]]
     41     By default, these cuts are off.  To use add to your command line -zerohalfCuts root (or other options) or just -zero.
     42     So far, they may help only on a small subset of problems and may need some tuning.[[BR]]
     43     The implementation of these cuts is described in[[BR]]
     44     G.{{{}}} Andreello, A. Caprara, and M. Fischetti[[BR]]
     45     "Embedding Cuts in a Branch and Cut Framework: a Computational Study with {0,1/2}-Cuts"[[BR]]
     46     INFORMS Journal on Computing 19(2), 229-238, 2007[[BR]]
     47     http://dx.doi.org/10.1287/ijoc.1050.0162
     48   * An alternative implementation of a **reduce and split cut generator** by Giacomo Nannicini is now available.
     49     By default, these cuts are off.  To use add to your command line -reduce2AndSplitCuts root (or other options).[[BR]]
     50     The implementation of these cuts is described in
     51     G.{{{}}} Cornuejols and G. Nannicini[[BR]]
     52     "Practical strategies for generating rank-1 split cuts in mixed-integer linear programming"[[BR]]
     53     Mathematical Programming Computation 3(4), 281-318, 2011[[BR]]
     54     http://dx.doi.org/10.1007/s12532-011-0028-6
     55   * An alternative robust implementation of a **Gomory cut generator** by Giacomo Nannicini is now available.[[BR]]
     56     By default, these cuts are off.  To use add to your command line -GMI root (or other options).[[BR]]
     57     The implementation of these cuts is described in[[BR]]
     58     G.{{{}}} Cornuejols, F. Margot, and G. Nannicini[[BR]]
     59     "On the safety of Gomory cut generators"[[BR]]
     60     http://faculty.sutd.edu.sg/~nannicini/index.php?page=publications
     61   * To encourage the use of some of the more exotic/expensive cut generators a **parameter -slowcutpasses** has been added.[[BR]]
     62     The idea is that the code does these cuts just a few times - less than the more usual cuts.  The default is 10.
     63     The cut generators identified by "may be slow" at present are just Lift and project and !ReduceAndSplit (both versions).
     64   * Allow **initialization of random seed** by user.  Pseudo-random numbers are used in Cbc and Clp.  In Clp they
     65     are used to break ties in degenerate problems, while in Cbc heuristics such as the Feasibility Pump use them   
     66     to decide whether to round up or down.  So if a different pseudo-random seed is given to Clp then you may get
     67     a different continuous optimum and so different cuts and heuristic solutions. This can be switched on by
     68     setting randomSeed for Clp and/or randomCbcSeed for Cbc.  The special value of 0 tells code to use time of day
     69     for initial seed.
     70   * Building on this idea, Andrea Lodi, Matteo Fischetti, Michele Monaci, Domenico Salvagnin, Yuji Shinano, and Andrea Tramontani
     71     suggest that this idea be improved by **running at the root node with multiple copies** of solver, each
     72     with its own different seed and then passing in the solutions and cuts so that the main solver has a richer
     73     set of solutions and possibly stronger cuts.  This is switched on by setting -multipleRootPasses.  These can also
     74     be done in parallel.
     75   * Few **changes to presolve** for special variables and badly scaled problems (in !CoinUtils).
     76   * New option -extraVariables <number> which switches on a trivial re-formulation that introduces extra integer variables
     77     to **group together variables with same cost**.
     78   * For some problems, cut generators and general branching work better if the problem would be infeasible if the cost is too high.
     79     If the new option -constraintFromCutoff is set, the **objective function is added as a constraint** which rhs is set to the current
     80     cutoff value (objective value of best known solution).
     81   * Click [https://projects.coin-or.org/Cbc/changeset?old_path=%2Freleases%2F2.7.8&new_path=%2Freleases%2F2.8.0 here] to see changes from release [https://projects.coin-or.org/Cbc/browser/releases/2.7.8 2.7.8]
    2482
    2583November 21, 2012
     
    111169
    112170You can obtain the Cbc source code either via subversion or in form of nightly generated tarballs.  The recommended method is to use subversion because it makes it easier to obtain updates. The following commands may be used to obtain and build Cbc from the source code using subversion:
    113  1. {{{svn co https://projects.coin-or.org/svn/Cbc/stable/2.7 coin-Cbc}}}
     171 1. {{{svn co https://projects.coin-or.org/svn/Cbc/stable/2.8 coin-Cbc}}}
    114172 1. {{{cd coin-Cbc}}}
    115173 1. {{{./configure -C}}}