Algorithm 801: POLSYS_PLP: A partitioned linear product homotopy code for solving polynomial systems of equations

Citation
Sm. Wise et al., Algorithm 801: POLSYS_PLP: A partitioned linear product homotopy code for solving polynomial systems of equations, ACM T MATH, 26(1), 2000, pp. 176-200
Citations number
34
Categorie Soggetti
Computer Science & Engineering
Journal title
ACM TRANSACTIONS ON MATHEMATICAL SOFTWARE
ISSN journal
00983500 → ACNP
Volume
26
Issue
1
Year of publication
2000
Pages
176 - 200
Database
ISI
SICI code
0098-3500(200003)26:1<176:A8PAPL>2.0.ZU;2-B
Abstract
Globally convergent, probability-one homotopy methods have proven to be ver y effective for finding all the isolated solutions to polynomial systems of equations. After many years of development, homotopy path trackers based o n probability-one homotopy methods are reliable and fast. Now, theoretical advances reducing the number of homotopy paths that must be tracked, and in the handling of singular solutions, have made probability-one homotopy met hods even more practical. POLSYS-PLP consists of Fortran 90 modules for fin ding all isolated solutions of a complex coefficient polynomial system of e quations. The package is intended to be used in conjunction with HOMPACK90 (Algorithm 777) and makos extensive use of Fortran 90 derived data types to support a partitioned linear product (PLP) polynomial system structure. PL P structure is a generalization of m-homogeneous structure, whereby each co mponent of the system can have a different m-homogeneous structure. The cod e requires a PLP structure as input, and although finding the optimal PLP s tructure is a difficult combinatorial problem, generally physical or engine ering intuition about a problem yields a very good PLP structure. POLSYS-PL P employs a sophisticated power series end game for handling singular solut ions, and provides support for problem definition both at a high level and via hand-crafted code. Different PLP structures and their corresponding Bez out numbers can be systematically explored before committing to root findin g.