ALGLIB started in 1999 and has a long history of steady development with roughly 1-3 releases per year. It is used by several open-source projects, commercial libraries, and applications (e.g.
TOL project,
Math.NET Numerics,[1][2]SpaceClaim[3]).
Features
Distinctive features of the library are:
Support for several programming languages with identical APIs (as of 2023, it supports C++, C#, FreePascal/Delphi, VB.NET, Python, and Java)
Self-contained code with no mandatory external dependencies and easy installation
Portability (it was tested under x86/x86-64/ARM,
Windows and
Linux)
Two independent backends (pure C# implementation, native C implementation) with automatically generated APIs (C++, C#, ...)
Same functionality of commercial and GPL versions, with enhancements for speed and parallelism provided in the commercial version
The most actively developed parts of ALGLIB are:
Linear algebra, offering a comprehensive set of both dense and sparse linear solvers and factorizations
Interpolation, featuring standard algorithms like polynomials and 1D/2D splines, as well as several unique large-scale interpolation/fitting algorithms. These include penalized 1D/2D splines, fast thin plate splines and fast polyharmonic splines, all scalable to hundreds of thousands of points.
Least squares solvers, including linear/nonlinear unconstrained and constrained least squares and curve fitting solvers
Optimization, with LP, QP and NLP solvers, derivative-free global solvers and multiobjective optimization algorithms.
Data analysis, with various algorithms being implemented