Set-based graphs (SB-Graphs) are graphs in which the vertices and edges are grouped in sets allowing sometimes a compact representation. When the compact representation is feasible, it is possible to develop new versions of traditional algorithms that take advantage of the structure of SB-Graphs, with a cost that is independent on the size of the sets of vertices and edges. Current algorithms for SB-Graphs:

• Connected Components
• Matching
• Strongly Connected Components (SCC)

This library defines data structures to represent SB-Graphs, and implements the aforementioned algorithms. The related publications can be used as documentation of the code.

This new approach was used in the flatter and causalization stage of the Modelica compiler, ModelicaCC (https://github.com/CIFASIS/modelicacc). Nevertheless, many fields could benefit of its use.

[1] Denise Marzorati, Joaquin Fernández, Ernesto Kofman. Connected Components in Undirected Set--Based Graphs. Applications in Object--Oriented Model Manipulation Applied Mathematics and Computation, Volume 418, 2022, 126842,ISSN 0096-3003, https://doi.org/10.1016/j.amc.2021.126842.

[2] Ernesto Kofman, Joaquín Fernández, Denise Marzorati. Compact sparse symbolic Jacobian computation in large systems of ODEs Applied Mathematics and Computation, Volume 403, 2021, 126181, ISSN 0096-3003, https://doi.org/10.1016/j.amc.2021.126181.

[3] Pablo Zimmermann, Joaquin Fernandez, Ernesto Kofman. Set-based graph methods for fast equation sorting in large DAE systems EOOLT '19: Proceedings of the 9th International Workshop on Equation-based Object-oriented Modeling Languages and Tools 2019

These are generic installation instructions.

In order to be able to install and compile SBG Library, the following dependencies must be installed:

• autoconf 2.69 (avoid 2.71)
• boost1.81 (or later)
• cmake
• g++
• make
• doxygen (optional)

The simplest way to compile this package is:

1. cd to the directory containing the package's source code and type autoconf to generate the configuration scripts.

2. Type ./configure to run the configuration script.

3. Type make to compile the library libsbgraph.a

4. Type sudo make install to install the library and header files in the default installation folders. The default installation folders are:

   • prefix=/usr/local
   • includedir=/usr/local/include
   • libdir=/usr/local/lib

5. You can remove the generated library and object files from the source code directory by typing make clean.

The makefile script accepts the following options:

• MODE = <Debug|Release> When set to Debug (default), adds the compiler's debug flags.

• prefix = Set the prefix installation path, default: /usr/local.

• exec_prefix = Set the binaries installation path, default: /usr/local.

• includedir = Set the header installation path, default: /usr/local/incldue.

• libdir = Set the library installation path, default: /usr/local/lib.

The makefile script accepts the following targets:

• test: Builds and run integration and unit tests.

• doc: Builds the documentation.

The SBG library is composed by four main modules:

• ast

• parser

• eval

• sbg

The first three were developed to allow for more user-friendly input. The first three were developed to allow for more user-friendly input, and the last one contains all the logical implementation of structures and operations.

Examples of SBG programs are presented in the /test folder with the '.test' extension. The bin/sbg-parser and bin/sbg-eval binaries accept these files as their input. Also, the make test command in the root directory will execute all of the existing tests.

Please see the file called LICENSE.

Report bugs to: [email protected] or [email protected]