!!! Development moved to GitLab !!!
Documentation to be found at: https://hzdr-fwdf.github.io/RISA/index.html
Prerequisites:
- cmake (>= 3.5)
- CUDA and compatible host compiler (e.g. gcc)
- libconfig++
- Boost (>= 1.58)
- LibTiff
Installation on Ubuntu 16.04
- install libConfig:
sudo apt-get install libconfig++-dev
- install cmake:
sudo apt-get install cmake cmake-curses-gui
- install boost:
sudo apt-get install libboost-all-dev
- install libtiff:
sudo apt-get install libtiff5-dev
- install CUDA (download from NVIDIA homepage: https://developer.nvidia.com/cuda-toolkit and follow the instructions)
- in order to build the documentation,
doxygen
andgraphviz
need to be installedsudo apt install doxygen graphviz
Building the software:
- If the software is built for another CUDA device architecture, this architecture needs to be added in the top-level CmakeLists.txt
- create a new path, outside of src-directory, called e.g. 'build'
- move into this path and enter following command in terminal:
ccmake <path_to_src>
e.g.ccmake ../RISA/.
- check if everything could be found and enter
CMAKE_BUILD_TYPE
, options are:Debug, RelWithDebInfo, Release
- if everything worked out, make the project
make -j all
- if build was successful, there is an executable in the
build/bin
folder - the documentation can be found in the
doc/html
folder
Running the software:
- first the configuration file in the example directory needs to be adapted. Especially the paths to the input data.
- the software can be invoked by issuing the following command:
./example <path_to_configuration_file>