Google Optimization Tools (OR-Tools) is a fast and portable software suite for solving combinatorial optimization problems.

• or-tools - Google's Operations Research tools in GitHub

  1. Install Anaconda Python 3.6 or 3.7
  2. Open a conda prompt
  3. Install or-tools 6.8 by running pip install ortools

SCIP is currently one of the fastest non-commercial solvers for mixed integer programming (MIP) and mixed integer nonlinear programming (MINLP). It is also a framework for constraint integer programming and branch-cut-and-price. It allows for total control of the solution process and the access of detailed information down to the guts of the solver.

• Installation on Mac OS X 10.13.6

  1. Download SCIPOptSuite-6.0.0-Darwin.dmg
  2. Open SCIPOptSuite-6.0.0-Darwin.dmg and copy the three directories, bin, include, and lib into some place (e.g., sudo mkdir -p /opt/scip and then sudo cp -R /Volumes/SCIPOptSuite-6.0.0-Darwin/bin /Volumes/SCIPOptSuite-6.0.0-Darwin/include /Volumes/SCIPOptSuite-6.0.0-Darwin/lib /opt/scip)
  3. Assume Anaconda Python is already installed on /opt/conda
  4. Change the dyld path as follows: (Repeat the same process for soplex and zimpl as well as /opt/scip/lib/libscip.6.0.0.0.dylib instead of scip.)

    $ cd /opt/scip/bin
    $ otool -L scip (shows the current dyld path)
    scip:
            /usr/local/opt/gmp/lib/libgmp.10.dylib (...)
            /usr/local/opt/gmp/lib/libgmpxx.4.dylib (...)
            ...
    $ sudo install_name_tool -change /usr/local/opt/gmp/lib/libgmp.10.dylib @rpath/libgmp.10.dylib scip
    $ sudo install_name_tool -change /usr/local/opt/gmp/lib/libgmpxx.4.dylib @rpath/libgmpxx.4.dylib scip
    $ sudo install_name_tool -add_rpath /opt/conda/lib
    $ otool -L scip (checks whether the dyld path was changed properly)
    scip:
            @rpath/libgmp.10.dylib (...)
            @rpath/libgmpxx.4.dylib (...)
            ...
    $ scip (checks whether it works)
    SCIP version 6.0.0 ...

• Installation on Windows Subsystem for Linux (e.g. openSUSE Leap 42 under Windows Subsystem for Linux)

  1. Download SCIPOptSuite-6.0.0-Linux.sh
  2. Install by running sudo bash SCIPOptSuite-6.0.0-Linux.sh --prefix=/opt and then cd /opt && sudo ln -s SCIPOptSuite-6.0.0-Linux scip
  3. Install the packages liblapack3 and libblas3 (in the case of openSUSE Leap 42, run sudo zypper install liblapack3 libblas3)

  Note: In the case of Cent OS 7, you may need patchelf to adjust the ld path. You can install it by running conda install patchelf under Linux.

• Installation on Windows 10

  1. Download SCIPOptSuite-6.0.0-win64-VS15.exe
  2. Install by double click the icon of the file SCIPOptSuite-6.0.0-win64-VS15.exe
  3. Be sure to add the path.

• PySCIPOpt - Python interface for the SCIP Optimization Suite

  1. Install Anaconda Python 3.6
  2. Install SCIP 6.0.0
  3. Open a conda prompt and run conda activate
  4. Check that cython is already installed by running conda list | grep cython. Otherwise, run conda install cython
  5. Set the environment variable export SCIPOPTDIR=/opt/scip (Linux & Mac OS X) or set SCIPOPTDIR=C:\Program Files\SCIPOptSuite 6.0.0 (Windows 10) (See INSTALL.rst for more details)
  6. Install PySCIPOpt 2.0.0 by running pip install pyscipopt
  7. How to check whether PySCIPOPt 2.0.0 is installed properly.
  • Mac OS X: DYLD_LIBRARY_PATH=/opt/scip/lib; ipython -c 'import pyscipopt' or export DYLD_LIBRARY_PATH=/opt/scip/lib
  • Linux: LD_LIBRARY_PATH=/opt/scip/lib; ipython -c 'import pyscipopt' or export LD_LIBRARY_PATH=/opt/scip/lib
  • Windows 10: ipython -c "import pyscipopt"
  8. (Optional: Mac OS X) Change the dyld path directly:

    $ cd /opt/conda/lib/python3.6/site-packages/pyscipopt
    $ otool -L scip.cpython-36m-darwin.so (shows the current dyld path)
    scip.cpython-36m-darwin.so:
	          libscip.6.0.dylib (...)
            ...
    $ sudo install_name_tool -change libscip.6.0.dylib @rpath/libscip.6.0.dylib -add_rpath /opt/scip/lib scip.cpython-36m-darwin.so
    $ otool -L scip.cpython-36m-darwin.so (checks whether the dyld path was changed properly)
    scip.cpython-36m-darwin.so:
	          @rpath/libscip.6.0.dylib (...)
            ...
    $ ipython -c 'import pyscipopt' (checks whether it works)