Hi folks,I'm interested in adding to my skills by learning a new scripting language. I know quite of bit of PHP and a bit of VBA, but I'm thinking about either Perl, Python, or Ruby. I'm tending to lean towards Python since it seems to have the best integration with C++ and is an touted as an "agile" language, whatever that means, but I'd like to get your thoughts. What language would you learn, and what's most used in the quant world?

I don't much about Ruby, but between PHP, VBA, Perl and Python, I would very much say Python is the most suitable for numerical work of any kind. I've been using C++ + SWIG + Python on and off for the last six years or so and it is by far the best environment I've worked in.

Folks in engineering are converging on python / scipy -- python seems the most vertile of those you mentioned. What do you mean "agile?"

QuoteOriginally posted by: TheDieselFolks in engineering are converging on python / scipy -- python seems the most vertile of those you mentioned. What do you mean "agile?"Google turned up thisBut I tend to think of an agile language as one that makes it easy to do Agile programming practices like refactoring and TDD. In practice, this mostly means that it supports reflection and is easy to parse so that refactoring and TDD tools can be built around it. Java might qualify, but I also tend to associate agile-related engineering practices with being "lightweight" so I feel high-level languages like Python or Ruby qualify a little more.C++ for instance would be the alternative: a language that makes it possible to write very fast and efficient code, but one that is not very agile.

I would say python ahead of php, perl or vba.I've used all four and python is:- easier to read - quicker to learn- has a better and more accessible standard library plus numerical add-ons (numpy, scipy)- has a good object orientated model if you are so inclined - integrates easily with Eclipse (with pydev)- is easy to integrate with C/C++- is fun!Enjoy,Rufus

I wouldn't pay much attention to the agile aspect of a language. Agile is, as far as I'm aware, only a software development process (we use it actually, it's quite good, I was skeptical at first).Saying C++ isn't agile is bollocks in reality.

QuoteOriginally posted by: dirtydroogI wouldn't pay much attention to the agile aspect of a language. Agile is, as far as I'm aware, only a software development process (we use it actually, it's quite good, I was skeptical at first).Saying C++ isn't agile is bollocks in reality.What kind of Agile software process do you guys use? Scrum, XP, Crystal Clear, something else?

QuoteOriginally posted by: ThomasJQuoteOriginally posted by: dirtydroogI wouldn't pay much attention to the agile aspect of a language. Agile is, as far as I'm aware, only a software development process (we use it actually, it's quite good, I was skeptical at first).Saying C++ isn't agile is bollocks in reality.What kind of Agile software process do you guys use? Scrum, XP, Crystal Clear, something else?Scrum, they thought about XP but I (am some others) refused. I couldn't share a keyboard and mouse with some unhygenic beast.It works quite well, it depends largely on the honesty of your developers but it soon becomes obvious who the weakest or laziest are.Project Managers have input at only certain points in the process which is really good, we actually had to tell one guy to f**k off as he waspestering us to do stuff that wasn't allotted for in that sprint, and we were well within our rights to do so.

python. it is very easy to learn and is a robust language with better OO than perl. it has a large library and user base (scipy, numpy, twistedmatrix, matplotlib, wx,...). It has a number of ways for extending in C: ctypes, swig, boost.python. mixing with R is trivial with rpy. And the ipython shell now has parallel support to enable distributed computations.

Either Ruby or Python. I personally prefer Ruby, but Python seems to have more momentum behind it at the moment.