Algebraic Approaches (TAPE)
- Goal: Use algebraic methods to model, reason about and compute CAs
- Idea: Translate coverage requirements of CAs to equation systems of multivariate polynomials
- Methodology: CAs arise as points in varieties when applying solvers relying on Gröbner Bases
- Real-World Challenges: Runtime of solvers
- Future Work:
- Analyse structure of equation systems for dedicated solvers
- Integrate applied modelling requirements for CAs (constraints, weights, etc.)
A Plug-in construction for CAs
- Goal: Construct CAs with more factors from CAs with less factors
- Idea: Adapt plug-in construction from classic design theory for CAs
- Methodology: Make use of coverage inheritance
- Application: Combinatorial Testing for contemporary composed Software Design
- Implementation: Haskell
Set-based appproaches (IFS)
- Goal: Generation of small CAs
- Idea: Enforce balancing properties to prune search space
- Methodology: Coverage property equivalent to intersection property of set systems
- Real-World Challenges: Runtime
- Implementation: Haskell
Fast In-Parameter-Order Algorithm - FIPO
- Focus: Efficient implementation of the IPOG algorithm
- Result: Orders of magnitude faster CA generation
- Optimization: Several improvements to existing algorithm
- Modelling: Two dimensional growth of CAs until complete
- Implementation: Rust
- CAgen: Publically available here
CAmetrics - Tool
- Motivation: Provide a tool for combinatorial analysis for test suites
- Advantages:
- Offers more features than existing tools
- Offers analysis tuned for performance or memory usage
- Web UI and command line interface
- CAmetrics: Publically available here
