autographix.ca

What It Is

AutoGraphiX (AGX) is a computational system for graph theory research, with the core goal of searching for extremal graphs: graphs that minimize or maximize a graph invariant, or a function of invariants, under given conditions. It is used not only to find candidate graphs, but also to extract information from extremal results, helping researchers automatically generate or manually discover new graph theory conjectures.

Core Capabilities

Based on the main text, AGX’s long-term development has focused on optimization modeling and ease of use. Early versions were mainly based on invariants, while AGX-2 introduced matrix and vector processing, allowing users to define new invariants without recompiling the software. AGX-III rewrote the interface and optimizer, adding vertex-based optimization and native multi-objective optimization. It can handle two objectives: either finding Pareto-optimal solutions, or setting a primary objective with a secondary objective used for tie-breaking. It also supports predefined constraints, such as maximum degree and diameter, reducing the burden on researchers to manually construct complex combinatorial optimization constraints. For special graph classes such as regular graphs, the system selects more suitable optimization routines and property-preserving transformations.

Pricing and Deployment

The main text only shows the current version as AGX-3.4.8 and provides version information for Mac OSX, Windows, and Linux. It does not state whether the software is free, commercially licensed, open source, or what payment methods are available. It appears more like local desktop/research software, but the text does not indicate whether it supports source compilation, self-hosted services, or batch-processing interfaces.

Pros and Cons

Its strengths are its highly specialized positioning, covering extremal graph search, user-defined invariants, multi-objective optimization, and conjecture discovery, making it suitable for graph theory and combinatorial optimization research. AGX-III abstracts constraints and optimization workflows, lowering the barrier to entry for users who are not combinatorial optimization experts. The downside is that the website provides limited public information and does not clarify its API/SDK, integration ecosystem, license, pricing, or maintenance support. For developers, its ability to support automated calls and integrate with existing research workflows is unclear.

Who It’s For and Access from China

It is best suited for researchers in graph theory, discrete mathematics, and combinatorial optimization at universities and research institutions, particularly for exploring extremal problems and forming conjectures. Access from mainland China cannot be determined from the main text and is marked as unknown.