arbitrary-order automatic differentiation library
libtnp offers an implementation for the AD method described in [1]
Requirements:
* C99 compiler
* C++11 compiler
* Boost http://www.boost.org/
* cmake
* make
If you are running on linux or a decent unix, you should be able to install all required dependencies using your package manager, e.g.
apt-get install build-essential cmake libboost-all-dev
Note, that both gcc and clang should work well.
In any other case (i.e. MS Windows) you will have to download and install all required tools yourself.
Since cmake favours out-of-tree builds, the next step is to create a dedicated build directory and run the build process in it.
mkdir build
cd build
cmake .. # add your specific cmake flags here
make && make install
In case of non-standard installations, specific installation targets etc. you have to set the appropriate cmake flags.
[1] Operational Semantics for a Modular Equation Language, C.Höger, 2013, Proceedings of the 4th Analytic Virtual Integration of Cyber-Physical Systems Workshop, December 3, Vancouver, Canada