English

The Lie algebraic structure of colored networks

Dynamical Systems 2024-09-24 v2

Abstract

In the computation of the normal form of a colored network vector field, following the semigroup(oid) approach in [19], one would like to be able to say something about the structure of the Lie algebra of linear colored network vector fields. Unlike the purely abstract approach in [10], we describe here a concrete algorithm that gives us the Levi decomposition. If we apply this algorithm to a given subalgebra, it does put the elements in the subalgebra in the block form given by the Levi decomposition, but this need not be the Levi decomposition of the given subalgebra. We show that for NN-dimensional vector fields with C colors (different functions describing different types of cells in the network) this Lie algebra netC,Nnet_{C,N} is isomorphic to the semidirect sum of a semisimple part, consisting of two simple components slC\mathfrak{sl}_C and slB\mathfrak{sl}_B, with B=NCB=N-C, which we write as a block-matrix and a solvable part, consisting of two elements representing the identity CC in cglCc\simeq glC and B in bglBb \simeq glB, and an abelian algebra aGr(C,N)a\simeq\mathfrak{Gr}(C,N), the Grassmannian, consisting of the CC-dimensional subspaces of RN\mathbb{R}^N. The methods in this paper can be immediately applied to study the linear maps of colored networks.

Keywords

Cite

@article{arxiv.2109.11419,
  title  = {The Lie algebraic structure of colored networks},
  author = {Fahimeh Mokhtari and Jan Sanders},
  journal= {arXiv preprint arXiv:2109.11419},
  year   = {2024}
}