From what I can gather this is a never-ending question. Narrowing it down a bit... can anyone make a recommendation (R vs Matlab) for someone heavily invested in Excel/VBA? In other words... which option will integrate better with Excel? on nuclearphynance: Matlab vs R vs Mathematica vs X vs Y vs .... http://www.nuclearphynance.com/Show Post.aspx?PostIDKey=138632 Matlab versus R http://www.nuclearphynance.com/Show Post.aspx?PostIDKey=91257 Matlab for Beginners http://www.nuclearphynance.com/Show Post.aspx?PostIDKey=132499 also found info on quant.stackexchange.com: What programming languages are most commonly used in quantitative finance? http://quant.stackexchange.com/ques...re-most-commonly-used-in-quantitative-finance side note: looks like R has some strong webapp capabilities http://yeroon.net/ggplot2/ also: Octave appears to be pretty popular as a basic/free MATLAB
SCILAB is the best MATLAB clone. It is also free, and both source code and binary are available. Documentation is excellent. Useless for my purposes, but there you go, for all you grail seekers... Other options are OCTAVE and RLAB, with OCTAVE the most compatible. Did I say that they are all free and open source?. Did I also say that they cost $0?. Very cheap and powerful, and they are also free. You don't even have to pay for them. All open source, and very cheap. Free actually.
I wanted to add to my previous post that the mentioned programs are all free, except MATLAB, which costs a fortune compared with the others, that are very inexpensive. Actually they are free. You pay nothing for them. If you want MATLAB, it can also be gotten for free but not legally, so I'm not interested really, when I can do (or not do) the same with SCILAB, which is free and legal, and open source, and can be easily translated to assembler at low cost. Yeah.
Another good feature of SCILAB is that the documentation is also free. It costs you $0. Like the actual program, source code included, which also costs $0. But you knew that already.
MATLAB is virtually free if you have a student license and if you can`t figure out how to buy it for the student price without a license, well, then...
LF: This time Sam beat you on the entry price ($0). But you and Sam (Sam because he discovered it, and you because he told you), already know that the exit (output) is much more important than the entry (input). It's nice to see that this alias is not the only addict to ET.
I have not used Matlab for a few years, what turned me off is the fact that they charge you for every little package. I have moved pretty much all of my alpha-generation code to R (i do primarily volatility arbitrage, with a little stat arb) already and I think will move most of my risk management code in the near future. The good: R has gone a long way over the past 3-4 years, it essentially went from a hobbyist project to a real programming environment. There are a few very good IDEs for R with encapsulated graphics, debugging and project management, the number of packages is staggering (a variety of packages for quantitative trading, for example). Also, it's all as free as it gets. The bad: R is not an intuitive language and even if you are a good programmer, it will take a while to learn. If you are more of a while-loop person, go for something more generic like python. Because of it's flexibility and "one-line"-ability, sometimes it is hard to read old code. Also, if you are writing non-interactive scripts, it takes some work to get it working properly.