Linear Algebra Functions in .NET (C#, Visual Basic)

ILNumerics supports the full set of basic linear algebra routines commonly needed in scientific algorithms. The collection contains all common matrix decompositions as listed below. Besides that, higher level functionality contains the solution of systems of Linear Equations and Eigenvalues / Eigenvectors.

• Linear Equation Systems, class reference: linsolve
• QR Decomposition, class reference: qr
• LU Decomposition, class reference: lu
• Cholesky Decomposition, class reference: chol
• Eigenvalue Decompositions, class reference: eig
• Pseudo Inverse of a Matrix, class reference: pinv
• Singular Value Decomposition, class reference: svd

Internal Implementation

Most of above functions internally utilize LAPACK routines from the official LAPACK maintenance repository. Since LAPACK is written in FORTRAN, most vendors have to include native libraries into their distribution. ILNumerics FORTRAN compiler translates LAPACK codes into efficient CIL code, based on ILNumerics smart arrays. Our purely managed LAPACK implementation runs on all .NET platforms. Support by native libraries is optional only.