Closed chobby closed 1 year ago
Hi @chobby, thanks for reaching out! Eigen and Boost.Hana are the only required dependencies to access the core functionalities of the library: these are demonstrated in the two examples variable
and variable_map
. Variables and variable maps can be integrated into any application where you need to organize all your data in a single vector.
The automatic differentiation and optimization capabilities are optional and can be enabled by setting the CMake flags UNGAR_ENABLE_AUTODIFF
and UNGAR_ENABLE_OPTIMIZATION
to ON
, respectively. For now, the latter option requires the former to be ON
, too—you will receive informative compile errors if not. When enabled, these options will automatically download the required dependencies via CMake. In particular, UNGAR_ENABLE_OPTIMIZATION
will download and make available OSQP and osqp-cpp. So, you do not have to install anything!
If you have a local installation of OSQP, then Ungar should be able to detect it by checking whether the target osqpstatic
is defined. As a final note, autodiff and optimization are only available on Linux operating systems because they rely on an autodiff library that uses features specific to Linux.
Thank you very much! I will give it a try.
Great, I am always available should you have any issues!
Thank you for developing a great library. The description only says that Eigen and Boost.Hana are required as dependencies, do I also need to install osqp further?