Latest release (Dec 2012): la4j-0.3.0.zip
la4j (Linear Algebra for Java) 
la4j - is open source and 100% Java library that provides Linear Algebra primitives and algorithms. There are the key features of la4j listed bellow:
- no dependencies and tiny size (~100kb jar)
- minimalistic and full object-oriented API
- sparse (CRS, CCS) and dense (1D/2D arrays) matrices
- linear systems solving (Gaussian, Jacobi, Zeidel, Square Root, Sweep and other)
- matrices decomposition (Eigenvalues, SVD, QR, LU, Cholesky and other)
- functors support: predicates, functions and procedures
- MatrixMarket/CSV IO formats support
Changelog
la4j-0.4.0 DEV. Planing to release at: Jun 2013
- added
rank()method (contributed by Evgenia Krivova); - fixed bug in
MatrixDecFunction; - added fast implementation of
determinant()method; - fixed bug in
MatrixMatketStream(contributed by Alessio Placitelli); - added
update()method (as compound operator replacement); - matrices are unsafe by default (new method pair -
safe()/unsafe()); - matrices and vectors support
slice()method; - support of
kronecker()product; - support map-reduce approach by method pair
transform()/fold(); - new matrices and vectors sources that handles IO streams;
- support of building a constant matrix in factories;
- matrices and vectors are immutable in terms of dimension;
- sparse entities are self-clearing now;
la4j-0.3.0 Dec 2012
- la4j got new API and new package naming (starting with "org.la4j.*")
- la4j has been moved from Google Code to GitHub;
- la4j uses Travis-CI;
- la4j available through Maven now;
- support of new matrices types (Basic1DMatrix, CCSMatrix);
- new I/O API and format - Symbol Separated Stream (CSV, TSV, etc.);
- new la4j entities: matrix/vector sources, matrix/vector functors;
- la4j uses default unchecked exceptions from java.lang.* now;
- the code has been formated according to JavaCodeConv;
- support of unsafe accessors and arithmetic methods;
- support Eigenvalues decomposition for non-symmetric matrix;
- new fast matrix-to-matrix multiply algorithm with blocks;
- new algorithm for runtime-based machine epsilon initialization;
- fixed several major/critical bugs;
la4j-0.2.0 Nov 2011
- new package la4j.io for reading/writing matrices in MatrixMarket format;
- matrices decomposition (LU, QR, Cholesky, SVD, Eigenvalues) support;
- matrix inversion support;
- sparse (Compressed Row Storage), dense matrices and vectors support;
- matrices and vectors can be serialized;
- la4j uses Maven and jUnit now;
la4j-0.1.0 Jan 2011
- eigenvalues decomposition;
- fast matrix multiply algorithm implementation;
la4j-0.0.7 Mar 2010
- solving linear systems;
- matrices transposing;
- finding of inverted matrix;
la4j 0.0.0 Jan 2010
- support real matrices and vectors;
Download
Details of the latest version of la4j can be found on the la4j project web site http://la4j.org or 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.
Contributors
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
by Vladimir Kostyukov, 2011-2013