la4j (Linear Algebra for Java)

The la4j is open source and 100% Java library that provides Linear Algebra primitives (matrices and vectors) and algorithms. The la4j was initially designed to be lightweight and simple tool for passionate Java developers. It has been started as student project and turned into one of the most popular Java packages for matrices and vectors.

The key features of the la4j are listed bellow:

Great performance allied with beautiful design
No dependencies and tiny size (~150kb jar)
Fluent and object-oriented/functional API
Sparse (CRS, CCS) and dense (1D/2D arrays) matrices and vectors
Linear systems solving (Gaussian, Jacobi, Zeidel, Square Root, Sweep and other)
Matrices decomposition (Eigenvalues/Eigenvectors, SVD, QR, LU, Cholesky and other)
Functors support: predicates, functions, procedures and accumulators
MatrixMarket/CSV IO formats support

Brief Examples

Matrix inversion

// We want simple dense matrix that uses 2D array as internal representation
Matrix a = new Basic2DMatrix(new double[][] {
   { 1.0, 2.0, 3.0 },
   { 4.0, 5.0, 6.0 },
   { 7.0, 8.0, 9.0 }
});

// We will use Gauss-Jordan method for inverting
MatrixInverter inverter = a.withInverter(LinearAlgebra.GAUSS_JORDAN);
// The 'b' matrix will be dense
Matrix b = inverter.invert(LinearAlgebra.DENSE_FACTORY);

System of linear equations

// The coefficient matrix 'a' is a CRS (Compressed Sparse Row) matrix 
Matrix a = new CRSMatrix(new double[][] {
   { 1.0, 2.0, 3.0 },
   { 4.0, 5.0, 6.0 },
   { 7.0, 8.0, 9.0 }
});

// A right hand side vector, which is simple dense vector
Vector b = new BasicVector(new double[] {
   1.0, 2.0, 3.0
});

// We will use standard Forward-Back Substitution method,
// which is based on LU decomposition and can be used with square systems
LinearSystemSolver solver = a.withSolver(LinearAlgebra.FORWARD_BACK_SUBSTITUTION);
// The 'x' vector will be sparse
Vector x = solver.solve(b, LinearAlgebra.SPARSE_FACTORY);

Matrix decomposition

// We want simple dense matrix, which is based on 1D double's array 
Matrix a = new Basic1DMatrix(new double[][] {
   { 1.0, 2.0, 3.0 },
   { 4.0, 5.0, 6.0 },
   { 7.0, 8.0, 9.0 }
});

// We will use LU decompositor
MatrixDecompositor decompositor = a.withDecompositor(LinearAlgebra.LU);
// The result should be treated as: L = lup[0], U = lup[1], P = lup[2]
Matrix[] lup = decompositor.decompose(LinearAlgebra.DENSE_FACTORY);

Download

Details of the last version of the la4j can be found on the project web site http://la4j.org or its GitHub page https://github.com/vkostyukov/la4j.

Licensing

This software is licensed under the terms you may find in the file named "LICENSE" in this directory.

How To Contribute

Fork it
Create your feature branch (git checkout -b my-new-feature)
Commit your changes (git commit -am 'Add some feature')
Push to the branch (git push origin my-new-feature)
Create new Pull Request

Contributors

The la4j wouldn't be the library it is today without the source contributions made by the authors:

Wajdy Essam
Evgenia Krivova
Julia Kostyukova
Alessio Placitelli
Pavel Kalaidin
Chandler May
Daniel Renshaw
Ewald Grusk
Jakob Moellers
Iurii Drozd
Maxim Samoylov
Phil Messenger
Anveshi Charuvaka
Clement Skau