Cylindrical decomposition for systems transcendental in the first variable (Q650841): Difference between revisions

The Cylindrical Algebraic Decomposition is a standard algorithm for solving problems in Real Algebraic Geometry by means of projection techniques. The algorithm deals with polynomial equations and inequalities over the real numbers. It has been recently extended (see [\textit{M. Achatz, S. McCallum} and \textit{V. Weispfenning}, in: ISSAC 2008. New York, NY: Association for Computing Machinery (ACM) 215--222 (2008; Zbl 1236.68301)]) to solve decision problems for systems involving exponential-polynomial in one of the variables and polynomial in the others. The paper under review is an extension of previous results by the author given in [\textit{A. Strzeboński}, in: ISSAC 2008. New York, NY: Association for Computing Machinery (ACM). 303--313 (2008; Zbl 1236.65057); in: ISSAC 2009. New York, NY: Association for Computing Machinery (ACM). 341--350 (2009; Zbl 1237.33015)]. It shows how to compute full cylindrical descriptions of solution sets of systems involving more general functions of the first variable. More precisely, it is shown how to compute algorithmically the cylindrical decomposition of the solution set as long as it is possible to solve univariate equations with functions from the algebra generated by the non-polynomial functions which appear in the input. Some implementations and experimental results are shown to illustrate the algorithm at the end of the text.