Real world comparison of Monte Carlo Gibbs simulations for HFT between C, C++, Java, Python, R, Scala, and Groovy? Surprised on the winner?
This is the closest real world comparison I could find. I think what is written is dead on! It seems Python with PyPy is a good choice but has little library support. Java is nearly as good as C so I have no point in relearning new languages like Groovy, Python, or Scala. They are not that popular in the banking world anyways so for career move that could be a time waster. There was a mention of using R for prototyping and extending through Java or C++ with R packages like Rcpp or RCppGSL. It looks impressive.
http://darrenjw.wordpress.com/2011/07/16/gibbs-sampler-in-various-languages-revisited/
The R with Rcpp and RcppGSL looks like promising too.
http://dirk.eddelbuettel.com/blog/2011/07/14/
I might go this route so we shall but posting definitely help out. Java with R is another very good option with Rcaller.
Get Matlab 2011a Coder Toolkit Generating Visual Studio 2010 Express DLLs With C++ Source Code
Are we speaking YOUR language?
Related posts:
- Why does Microsoft .NET project always trump over R, Perl, Java, and even Python?
- Youtube video: Online Course on open source C++ quant library QuantLib with QuantLibXL, Excel, Java, C#, R, Python
- Intro to QuantLib C++ free open source library for quants with Excel and Java, Python, C#, R support
- “Implementing QuantLib” drafts for Monte Carlo framework for quant development simulations
- Do you use NAG in Java for quant analaytics? Check out a comparison of the Java Native Interface (JNI) vs. Java Native Access (JNA)


