QCMatPack for Java Matrix and Numerical Analysis Software
The QCMatPack software implements what we consider the most useful numerical algorithms, written entirely in Java. That includes routines for the generalized solution of linear algebra problems, and specific algorithms for solutions of linear equations, eigensystems, digital signal processing, multiple regression, and curve fitting. Since the algorithms are implemented using 100% managed code, they are the better choice for workstation and web based Java applications.
QCMatPack is 100% managed code, using only the Java core numeric routines in the matrix and numerical analysis classes. The software makes NO calls to external, non-Java based libraries.
Simple to use, object oriented, matrix and numerical analysis classes for the most common problems in science, engineering, industry, and business.
There are no limits regarding the size of vectors and matrices, other than the amount of free memory on the host computer. There are no limits regarding the size of the vectors and matrices used in the solution of general linear systems, solution of eigenvalue problems, multiple regression, curve fitting and DSP routines.
Java IDE Independent - Tested and compatible with Eclipse, JBuilder and NetBeans.
Windows 32-bit/64-bit Compatible - The software will run as 32-bit on Windows 32-bit operating systems (XP and 32-bit versions of Vista and Windows 7) and as 64-bit on Windows 64-bit operating systems (64-bit versions of Vista and Windows 7). This is a feature of Java programs in general.
Complex number math - implements a complex number class, DComplex, that provides a data structure and functions for manipulating complex numbers.
Java based matrix math and linear algebra - The software utilizes several generalized matrix classes, the most important of which are the DDMat for real number matrices and the DXMat class for complex number matrices. These are used for the input and output of data to the numerical analysis routines and the manipulation of real and complex number matrices. The matrix classes are designed so that there is no practical limit, other than the amount of free system memory, to the size of matrices. The DDMat and DXMat are symmetrical, in that the same matrix math and linear algebra routines are available for both real and complex numbers. The matrix classes includes routines for matrix and vector statistics (mean, maximum, minimum, range, sum, variance, standard deviation, sum of squares), matrix norms (1-norm, 2-norm, infinity-norm), utility routines (saxpy, dot product), file I/O, general matrix math functions (addition, subtractions, multiplications, division, transpose, inverse, determinant) and matrix row/column manipulation (insert, delete, copy, and interchange).
Ax = b
Solving for eigenvalues and eigenvectors (Real and Complex)
AX = λX
QL, QR and LR algorithms
