This survey gives an overview of three central algebraic themes related to the study of splines: duality, group actions, and homology. Splines are piecewise polynomial functions of a prescribed order of smoothness on some subdivided domain D in R^k, and appear in applications ranging from approximation theory to geometric modeling to numerical analysis. Alternatively, splines can be interpreted as a collection of polynomials labeling the vertices of a (combinatorial) graph, with adjacent vertex-labels differing by a power of an affine linear form attached to the edge. In most cases of interest, the subdivided domain is essentially dual to the combinatorial graph, and these two characterizations of splines coincide. Properties of splines depend on combinatorics, topology, geometry, and symmetry of a simplicial or polyhedral subdivision of a region D in R^k, and are often quite subtle. We describe how duality, group actions, and homology -- techniques which play a central role in many areas of both pure and applied mathematics -- can be used to illuminate different questions about splines. Our target audience is nonspecialists: we provide a concrete introduction to these methods, and illustrate them with many examples in the context of splines. We also provide a tutorial on computational aspects: all of the objects appearing in this note may be studied using open source computer algebra software.
翻译:暂无翻译