Download

RobOptim is currently available on Linux and Mac OS X. Windows support is still experimental.

How to install RobOptim on your machine?

Choose your distribution in the list or choose source install if your distribution is unsupported. Instructions are for the Core package but other packages can be installed following the same steps, simply replace "core" by the appropiate package you want to install.

Package	Latest Release
RobOptim Core	3.2
RobOptim Core Python ^alpha	3.2
RobOptim Core Plugin CMinPack ^beta	3.2
RobOptim Core Plugin IPOPT	3.2
RobOptim Core Plugin Eigen ^beta	3.2
RobOptim Core Plugin NAG ^beta	n/a
RobOptim Core Plugin CMA-ES ^alpha	n/a
RobOptim Core Plugin NLopt ^beta	3.2
RobOptim Core Plugin PaGMO ^alpha	3.0
RobOptim Core Plugin MOSEK ^alpha	n/a
RobOptim Core Plugin KNITRO ^alpha	n/a
RobOptim Core Manifold ^beta	3.2
RobOptim Analysis ^beta	n/a
RobOptim Capsule	3.2
RobOptim Trajectory	3.2

While not available in the official repositories, RobOptim is available on Arch Linux using the AUR.

If you are using Arch Linux, please open a terminal and run the following command:

    # To install RobOptim Core with yaourt:
    $ yaourt roboptim-core
    # or with packer:
    $ packer roboptim-core

If you are using Debian or Ubuntu, please open a terminal and run the following command:

    $ sudo apt-get install libroboptim-core-dev

If you are using Fedora, please open a terminal and run the following command:

    $ sudo yum install roboptim-core-devel

If you are using RobotPkg, please open a terminal and run the following command:

    # from your robotpkg directory
    $ cd optimization/roboptim-core && make update

/!\ DISCLAIMER /!\ Windows/Visual Studio support is still experimental, so avoid it if you are not an experienced developer.

If you are using Windows, you will first need to install RobOptim's dependencies:

Git (available with chocolatey),
CMake (available with chocolatey),
Boost (from source),
doxygen (available with chocolatey),
ltdl (from source),
patched log4cxx (from source).

Your environment will also need to be set accordingly, with the following environment variables:

PKG_CONFIG_PATH: path to the pkgconfig directory containing .pc files (e.g. C:/devel/lib/pkgconfig).

You can finally start building RobOptim:

Clone the repository (e.g. https://github.com/roboptim/roboptim-core.git) and initialize/update its submodules,
Generate the Visual Studio solution with CMake (you will probably need to provide some CMake variables such as BOOST_ROOT and BOOST_LIBRARYDIR),
Compile the project and run tests in Visual Studio to make sure dependencies were properly installed.

RobOptim is available through an additional repository homebrew-roboptim.

If you are using Homebrew on Mac OS X, you will need to run the commands:

    # We make assumption that you have already installed Homebrew!
    # If this is not the case, please look at: http://brew.sh/

    # Add the repositories
    $ brew tap homebrew/python   # only for roboptim-core-python
    $ brew tap homebrew/science  # for IPOPT, CMinPack
    $ brew tap roboptim/roboptim

    # RobOptim Core
    $ brew install roboptim/roboptim/roboptim-core

    # RobOptim Core Plugin CMinPack
    $ brew install roboptim/roboptim/roboptim-core-plugin-cminpack

    # RobOptim Core Plugin IPOPT
    $ brew install roboptim/roboptim/roboptim-core-plugin-ipopt

    # RobOptim Trajectory
    $ brew install roboptim/roboptim/roboptim-trajectory

    # RobOptim Capsule
    $ brew install roboptim/roboptim/roboptim-capsule

    # IMPORTANT: some of these repositories are not released.
    # In this case, you can install from Git using:
    # brew install --HEAD a_package
    #
    # Note that in this case, you have to install *all* packages
    # from source (even roboptim-core!).

Compiling from source is significantly more challenging than installing a binary package. Be prepared to face compilation errors, software bugs and crashes.

Dependencies

First, install the following tools on your system:

CMake
gcc, g++
pkg-config
Doxygen
Boost
Eigen (≥ 3.2.0)
Log4cxx (≥ 0.10.0)
Libtool and its ltdl library
Optional: git (if not using a release tarball)

And then follow one of the two next steps depending on your needs.

Installing a release from source

    # Uncompress the tarball (replace X.Y by the proper version number)
    $ tar xzvf roboptim-core-X.Y.tar.gz
    # Move to the newly created directory
    $ cd roboptim-core-X.Y
    # Create build directory
    $ mkdir _build && cd _build
    # Run cmake or ccmake and set the installation options as needed
    # Note: RelWithDebInfo is the prefered value for CMAKE_BUILD_TYPE
    #       Debug (or no value) is considerable slower than release mode.
    $ cmake -DCMAKE_BUILD_TYPE=RelWithDebInfo -DCMAKE_INSTALL_PREFIX=/your/install/prefix ..
    # Build the software
    $ make
    # Run the test suite
    $ make test
    # Install the software
    $ make install

Installing from Git

    $ git clone --recursive https://github.com/roboptim/roboptim-core.git
    # Move to the newly created directory
    $ cd roboptim-core
    # Create build directory
    $ mkdir _build && cd _build
    # Run cmake or ccmake and set the installation options as needed
    # Note: RelWithDebInfo is the prefered value for CMAKE_BUILD_TYPE
    #       Debug (or no value) is considerable slower than release mode.
    $ cmake -DCMAKE_BUILD_TYPE=RelWithDebInfo -DCMAKE_INSTALL_PREFIX=/your/install/prefix ..
    # Build the software
    $ make
    # Run the test suite
    $ make test
    # Install the software
    $ make install