Not sure this is the right forum, but here goes: Let's say ssomeone with no programming experience wants to begin learning how to construct quantitative models for trading. What wolud be a good progression in learning programming languages? How about starting with something like VisualBasic.net, then moving up the ladder to more sophistocated prgramming? Where would you go next? C++? Java? My questions basically are: Which languages are best to use, initally to learn, then ultimately for the greatest versatility? Which prograssion from a simple to more complex language makes the most sense? Thanks for any help!
this has been discussed before. for an analyst/trader/researcher you can get by wit a scripting language ... perl python R matlab i would recommend python with the scipy module and quantlib module. you will everything you need to price, construct, analyze.
I started with python, then C and C++. I can also vouch for python as not only being the best introductory language, but one that is also powerful enough for very complex applications. Plenty has been already said here on ET.
Here is my question , What are you trying to accomplish and what type of educational background do you have ?
It depends on how complicated the model you try to create. If your models are based on technical indicators, EasyLanguage in Tradestation can do the job. As the name implied, EasyLanguage is easy to learn and to use. I'm not a programmer myself and I just spent a weekend to look through the sample codes to figure out what they mean, and I was able to code couple systems that week. Tradestation also provides a huge database of historical data for different financial instruments: equities, futures, etc. It also has the optimisation function which allows you to find the "best" parameters for your setup. For beginner, I recommand this program. However, Tradestation seems to suck as broker, maybe it's good to keep it as a tool for modeling strategies, but not as a transaction platform. I'm not sure though, I never traded with it. If your model are more complicated, implying fundemental inputs, then you need somehting more sophiscated than EasyLangague. Couple websites offer service for backtesting strategies based on fundamental. Good luck! (Oops, I didn't read through your entire post before I start posting, I just feel like to talk today, lol, I didn't know your focus is on coding languages instead of where to start off with modeling)
Ask yourself what kind of things you would like to achieve. Next look at solutions that are already on the market. If there is something valueable that no software out of the box can do for you than you have a reason to do it yourself. It takes quite some time to get proficient in programming. Don't get deluded by books like "Learning C++ in 21 days". Just my 2 cents. TM
http://freescienceonline.blogspot.com/2006/06/free-computer-science-video-lecture_24.html Go through "Structure and intrepretation of computer programs," "Programming Languages," "Introduction to Algorithms" and "Algorithms"and anything else that looks useful... then just pickup whatever language you want to learn the syntax for.
Well if u want to start learning a programming language using a general purpose language like VB.NET is a good start and end. There's no need to "upgrade" to the other .NET languages since they all compile to the same machine code. Unless you need super fast code, then you'll have to switch to C++ non .NET. But you may want to look at languages that can do operations on scalars which makes analyzing data easier like Tradestation or various other scripting languages available that can do the trick. I think Modulusfe has a component, a bit pricey. Sticking with the general purpose language will ultimately give you all the freedom and versatility you want, you'll have complete control. But it'll be more work. If you go the VB.NET route I'd strongly recommend also downloading the free SQL server edition to store and do some scarlar operations on the data.