BOLeRo is developed for the latest Ubuntu LTS (currently 16.04), has been tested with 17.04 and the current version of 18.04 and its continuous integration runs on 14.04. Using it on Mac OS or Windows is possible, however, the installation might be more complicated. We recommend to use Docker if you want to take a quick look at the software. Instructions are here.
In order to install BOLeRo and its dependencies from source, the pybob build system is used. Therefore, the operating system needs to be prepared according to the prepare sections in the documentation. Afterwards, the following installation instructions can be used.
There are two alternatives to install BOLeRo.
This is the recommended way to install BOLeRo. The installation process for BOLeRo is a little bit unusual because we depend on the latest versions of many libraries and support Linux, MacOS and Windows so that we usually have to download and build them. You can download the installation script and run it with
wget https://raw.githubusercontent.com/rock-learning/bolero/master/bootstrap_bolero.sh
chmod +x bootstrap_bolero.sh
./bootstrap_bolero.sh
The latest release of BOLeRo is also available at mloss.org or you can download it from Github. After you unzipped the release, you can build BOLeRo with the script install.sh from the folder bolero-dev. Note that you have to run it from that folder, otherwise the environment variables will not be configured correctly. The installation script will also install dependencies.
The installation script downloads all required packages and installs them in a target directory that you defined. We call it $BOLEROPATH. It has the following structure:
$BOLEROPATH
|-bolero-dev/
|--install/
|---bin/
|---configuration/
|---include/
|---lib/
|---share/
|--autoproj/
|---manifest
|--learning/
|---bolero/
|--env.sh
The subdirectory install includes all the shared libraries, configurations, header files, and scripts that you installed. Usually they are not available in your environment. However, you can source the script env.sh with
source env.sh
to make them available. You could add it to your .bashrc to make it permanent.
The subdirectory autoproj contains a file manifest which includes all activates packages of BOLeRo. You can activate commented packages by removing “# ” in front of it and calling bob-bootstrap. The manifest file follows the conventions of autoproj.
It is possible to install only the Python library BOLeRo without any C++ module via
python setup.py install
To run the Python unit tests, we need nosetests. You can install it with
sudo pip install nose
and run it with
nosetests bolero -sv
in the bolero main directory bolero-dev/learning/bolero.
Install dependencies:
sudo apt-get install doxygen
sudo pip install joblib pillow
Go to the folder ‘doc’ and run
make html
The result will be located in doc/build/html/index.html.