MPSolve

MPSolve
Name	MPSolve
Genre	Numerical analysis software

MPSolve

MPSolve is a specialized numerical software package for computing approximations to roots of univariate polynomials. It is used in research and engineering contexts that require high-precision rootfinding, combining certified algorithms with practical speed for polynomials arising in algebraic geometry, control theory, and scientific computing.

Overview

MPSolve implements algorithms for isolating and refining complex roots of polynomials with coefficients in fields such as Rational number, Floating-point arithmetic contexts or exact Integer (number), and it emphasizes certified output and arbitrary precision. The project interfaces with mathematical ecosystems such as SageMath, MATLAB, GNU Scientific Library, PARI/GP and tools in the SymPy and NumPy ecosystems, while its design reflects influences from researchers associated with Georgia Institute of Technology, Dartmouth College, and laboratories such as National Institute of Standards and Technology. Users range from contributors to Numerical Recipes style projects to teams involved with ISO numerical standards and mathematical software projects like Maple and Mathematica.

History and Development

Development traces to research efforts in polynomial rootfinding and validated numerics by authors who published in venues including the SIAM Journal on Numerical Analysis, ACM conferences, and proceedings of the International Congress of Mathematicians. Early prototypes drew on techniques studied by mathematicians associated with Princeton University, University of California, Berkeley, and ETH Zurich. Subsequent releases integrated advances from groups at University of Waterloo and École Polytechnique Fédérale de Lausanne that focused on multiprecision arithmetic and complex analysis methods. Contributions and comparisons have appeared alongside work from teams at Cornell University, University of Cambridge, and Massachusetts Institute of Technology.

Algorithms and Features

MPSolve combines subdivisions and Newton-type methods adapted to complex plane topology, leveraging multiprecision libraries inspired by GNU MP and algorithmic paradigms discussed in literature from Stanford University and California Institute of Technology. It uses certified root isolation techniques related to Pellet's theorem and Graeffe iterations, refined by near-optimal complexity bounds studied by researchers at University of Illinois Urbana–Champaign and University of Toronto. Implementation includes adaptive precision control influenced by designs from University of Oxford numerical groups and backward/forward error analysis approaches common in work by Argonne National Laboratory and Los Alamos National Laboratory. The software supports arbitrary-precision arithmetic, deflation strategies researched at Imperial College London, and error certification frameworks discussed in publications from Duke University and University of Michigan.

Usage and Interfaces

MPSolve provides command-line interfaces and libraries for integration with numerical stacks used at institutions such as CERN, NASA, and European Space Agency. Bindings and example wrappers have been developed for environments like Python (programming language), R (programming language), and Julia (programming language), making it accessible to practitioners affiliated with projects at Broad Institute and computational groups at Harvard University. The package is invoked in reproducible workflows alongside systems such as GitHub repositories maintained by research labs at Brown University and University of Pennsylvania, and it is used in coursework and projects related to curricula from MIT OpenCourseWare and Coursera offerings coauthored with faculty from Yale University.

Performance and Benchmarks

Benchmarking work compares MPSolve against tools like rootfinders embedded in Maple, Mathematica, and specialized libraries developed at NAG (Numerical Algorithms Group) and academic groups at Tokyo Institute of Technology. Reports in journals associated with Springer and Elsevier show that for dense and clustered-root polynomials MPSolve's multiprecision and certification strategies often outperform naive Newton iterations described in texts from Wiley and Cambridge University Press. Performance studies conducted in collaboration with teams at ETH Zurich and University of Bonn evaluate speed, robustness, and precision on testbeds drawn from competitions and datasets curated by International Mathematical Union related projects.

Applications and Examples

MPSolve is applied in computational algebraic geometry problems studied at Institut des Hautes Études Scientifiques and in control-systems pole-placement analyses from research groups at Siemens and Honeywell. It is used to validate eigenvalue computations in finite-element analyses carried out by teams at Argonne National Laboratory and to compute root-loci for circuits in collaboration with researchers at Bell Labs and IBM Research. Examples include certification of algebraic curve intersections in projects led by groups at Max Planck Institute for Mathematics and trajectory polynomial root analysis in aerospace studies at Jet Propulsion Laboratory.

Category:Mathematical software Category:Numerical analysis