Recommendation for C # matrix library

Question

I need a C # library to work with matrices. It should implement the decomposition of singular values, matrix inversion, etc.

I used to use CSML , but this does not implement some functions that I need.

Some of the features are:
Singular value decomposition
Matrix inversion
Trace
Determinants

Recommended libraries (from answers)

(most of them implement the functions indicated above)

• Math.NET Numerics for .NET, Silverlight, and Mono (open source)
• Extreme Optimization Numerical Libraries (commercial, very complete)
• IL Numerics (commercial)
• Deltix FinMath Digital Library (commercial, uses Intel® Mathematical Core Library (MKL) and Intel® Integrated Performance Primitives (IPP))
• Easy fast matrix class in C # (Strassen algorithm, LU decomposition)

Answer 1

Math.NET Numerics is very good if it supports the operations you want. Senior Math.Net Iridium still supports more features. In addition, dnAnalytics is pretty nice, but not developed anymore. (He, like Iridium, integrates into Math.NET Numerics.)

On the commercial side, there are very good, reliable options. Extreme Optimization Numerical Libraries work very well. The Visual Numerics library also works very well (albeit with royalty-based distribution).

Answer 2

I believe that Math.NET has most of the features you mention. Yuu may want to take a look at it and see if it suits your needs.

You can look at the Iridium samples to see the syntax and sample code.

Answer 3

I often used ILNumerics . He also made some contributions to the project. I know this, this is the only library that I know that does not spend half the processor time in the GC for serious computing . Its fast and mature. Full list .

Answer 4

In addition to the libraries mentioned in other posts, I can say a few kind words about the FinMath library. It is rather a new library, but, in my experience, it has several advantages. It is very simple to use this library. Firstly, it has only one executable file (* .dll) for 32-bit and 64-bit operating systems. Secondly, it is built on the basis of the well-known Intel Math Kernel Library and, therefore, provides very good performance. The managed .NET API provided by this library is also very extensive and flexible. But this is proprietary software and costs a significant amount of money. It also does not have some functionality (for example, compared to NMath from Software CenterSpace). In any case, it's worth a try.