The current project exposes interfaces in different languages to a pragmatic and efficient geodesy toolkit. Both spherical and WGS84 models are supported, though there are more functions implemented for the spherical model.
The source code is common to all languages and implemented with C++ templates for simple and double precision floating point values. When possible, SSE2 vectorisation is provided. (AVX expected) Interfaces are then provided for Python (through cython) and for OCaml.
The src folder contains the C++ implementation. Simply include resp. spherical_geodesy.h or wgs84_geodesy.h according to your needs. *.cpp files are just provided for aesthetical reasons, if you want to compile the library in advance.
The geodesy folder contains the Python interface and exposes both geodesy.sphere and geodesy.wgs84 modules. You can install the module with the setup.py file provided. numpy is necessary to compile the library.
Recommended installation procedure:
virtualenv (optional, but a good idea nonetheless):
virtualenv geo_test
source geo_test/bin/activatepip install cython numpy
pip install git+git://github.com/xolive/geodesy.gitYou can then read the manual embedded in geodesy.sphere and geodesy.wgs84
The ocaml folder contains the OCaml interface. Refer to geodesy.mli for the interface.
Recommended installation procedure using ocamlfind and opam (from version 1.2), provided your PATH yields access to cmake:
opam pin add geodesy git://github.com/xoolive/geodesy.gitYou can then compile your OCaml source file with ocamlfind, e.g.:
ocamlfind ocamlc -linkpkg -package oUnit,geodesy tests/test_geodesy.mlThe CMakeLists.txt also offers a decent installation procedure provided you want to embed the library in bigger projects. You will need to clone the ocaml-cmake project first and set the CMAKE_MODULE_PATH accordingly.
You can also use the make install target for a project installation that you may later import in a different project. (find_package (geodesy REQUIRED) in your new CMakeLists.txt)