A more efficient TMR implementation for large computations
- Boost 1.45
- Clang 8 or higher
- CMake 3.18 or higher
- Ninja 1.11.0 or higher
- Run
make build
in the root directory of the repo, which will automatically run CMake and Ninja - Output files will be in the
build/
subdirectory
For ease of use, a Docker image with all dependencies preinstalled is also available. The image can be built with:
make docker
A container can be started that maps the repository to /mnt
using:
make docker-shell
typedef pair<size_t, void *> DTSSInput;
struct InputData {
vector<DTSSInput> inputs;
DTSSInput output;
bool operator==(InputData& b);
};
dtss_compute(unordered_set<InputData *> dataset,
void (*processor)(InputData *));