Installation#

Pip#

The preferred way to install the alpaqa Python interface is using pip:

python3 -m pip install --upgrade --pre alpaqa

To compile problems using the Python interface, you will need a C compiler, such as GCC or Clang on Linux and MSVC on Windows.

Building alpaqa from source requires the installation of some C++ dependencies, see Installation (Doxygen) for detailed instructions.

Pre-built binaries for Linux are available from the Releases page on GitHub.

For Debian-based systems, the .deb packages can be installed using

sudo apt update
sudo apt install ./libalpaqa*_1.0.0a10_amd64.deb

Different components are available:

libalpaqa contains the shared libraries needed to run applications that use alpaqa.
libalpaqa-debug contains the debugging symbols for those libraries.
libalpaqa-dl_dev contains the C header files needed to compile problem specifications that can be dynamically loaded by alpaqa.
libalpaqa-dev contains all development files such as headers and CMake configuration files needed to compile software that invokes alpaqa solvers.
libalpaqa-tools contains command line utilities such as alpaqa-driver, which can be used to invoke the solvers directly, without the need to write any C++ code.

The following distributions are tested:

Alternatively, the .tar.gz file can be extracted and installed manually.

sudo tar xzf alpaqa-1.0.0a10-Linux-x86_64.tar.gz -C /usr/local --strip-components=1

This requires glibc 2.17 or later. You may need to install or pre-load the following additional runtime dependencies: