Open Source / Free Statistical Engine for .NET / C # Projects?

Question

I have an ASP.NET C # web application that requires some aggregate functions. Until now, I myself have written them myself, however, since the statistical side is expanding, I would prefer to reuse the open or free statistical engine as a library.

I looked at R and its DCOM interface (http://sunsite.univie.ac.at/rcom/). Although R is excellent, the DCOM interface is third-party and has restrictions on commercial use (I may have ads on the site). Not to mention that the DCOM interface is a bit slow. So this will not work for me.

Does anyone have any other methods or suggestions on what I'm trying to do?

Thanks!

Answer 1

Thanks grapeot. I also searched and found R.NET at http://rdotnet.codeplex.com/ . The source is also available using Hg / Mercurial (instead of SVN or GIT) via

hg clone https://hg01.codeplex.com/rdotnet 

I just tried this and it seems to be pretty decent for toy examples. I will have to test some larger examples in the real world to see its strength. In any case, with available source code, it should be easier to fix / contribute to the project :)

Some additional answers from another post:

Pat . While there is more to the ML engine and perhaps a way you don't need, you recently released Infer.NET. a> which can be a step in the right line?

Michelle : Have you looked at Python ? I know some people use this, for example. to increase the functionality of SPSS above its macro capabilities, but I don’t know what is available as complete functions.

Josh Hermann : Numpy and Scipy were ported to .NET last year, so you can use this functionality with IronPython. You will find many examples on the Internet where people mix C # and IronPython (although you can run CLR / C # vs DLR / IronPython)

Answer 2

You can try Accord.NET , a statistics and machine learning library in .NET. Personally, I use it in scientific computing and feel it expressive and quite effective.