遇见数据集

Replication Data for: A general mocked interface for the coupling library preCICE

收藏
Zenodo2026-01-19 更新2026-05-26 收录
官方服务:

资源简介:

This is the replication data for the bachelor's thesis "A general mocked interface for the coupling library preCICE" by Julian Harrison-Wirth, written at the IPVS of the University of Stuttgart. It contains all the scripts used in development and the results used in the thesis. Files:Developments_Scripts contains scripts used in development, mainly for compatibility and regression tests, and example outputs of the scripts Determinism_Test_Results contains a script used to test the repeatability of the mock's data generationProfiling_Scripts contains scripts to profile and benchmark runs of tutorials against the mock Benchmark_Results contains the results of two runs of the above script used to measure the runtime of a tutorial using the mock Thesis abstract: Testing adapters for the preCICE coupling library requires full solver installations and complete coupling workflows, resulting in slow feedback cycles, brittle test environments, and high development overhead. This thesis presents a lightweight mock implementation of the preCICE Participant API that addresses these challenges through configuration parsing, deterministic data modes, and simulation of key lifecycle behaviors. The mock is implemented as a single C++ source file with minimal dependencies and integrated into existing preCICE build systems. The mock enables adapter testing in isolation without external solver dependencies, provides fully reproducible outputs through seeded pseudo-random data generation, and supports both explicit and implicit coupling schemes. Empirical validation demonstrates API compatibility with existing preCICE tutorials and training materials, deterministic behavior across repeated runs, and 2-5$\times$ performance improvements over full coupled simulations. By eliminating most external dependencies and enabling fast, deterministic test execution, the mock facilitates test-driven development practices and integration into continuous integration pipelines.

提供机构:
Zenodo
创建时间:
2026-01-19
二维码
社区交流群
二维码
科研交流群
商业服务