Anybody have any experience? I'm looking to expand my knowledge professionally by picking this up in addition to SQL and wanted to know who here is the guru. Any advice, resources, etc?
Learned it in college. Very easy to use and there are extensive tutorials on how to use this software. Can be a pain at times though.
Were you hoping Daryl would respond by posting it in the GARM? :grin: I thought about playing around with it just for fun, nothing serious, but have way too much junk I'm trying to do now. I'd just Google for forums, beginner's sites, or buy a book or two on it. Out of curiosity, what do you plan on doing with it?
...mehbeh.... My curiosity isn't far from where you were. The ability to represent big data in different ways could benefit me professionally, but honestly, i just enjoy working with data sets. I've figured out how to scrape NBA stats into my personal tables. Now just goofing around with it. What would you guys consider alternatives to R?
If it's for professional use, SPSS or SAS (especially if working with big data sets) if you want a "pure" statistical software. Though obviously these options aren't viable if you *have to* pay for them because they're very expensive. Otherwise, if you're just working with NBA stats and wanna mess around, might as well use R because it's free. The learning curve isn't that bad. And it'll get what you need done. Keep in mind that many programming languages (if you already know one) have lots of awesome statistical/math packages that you can use, i.e. python.
I saw this book at Barnes and Nobles this evening. Do you have $22? Because that is how much it costs: $22. This course includes a section on R programming, and includes a certificate upon successful completion. It is created by Johns Hopkins University and costs a few hundred $. Or you can pay more for a program like this run by the University of Washington and costs several thousand $.
If you decide to go with R. Or if you decide to learn a more complete common language that will allow you to use a great statistical package, such as c++, Python, etc... A good idea might be to take a free course from "Coursera" or "edx" I've used coursera to brush up on some financial math (for options) before and it was well done. They definitely should have a lot of coding courses that'll help you get used to using something like R without having to do a brunt of the information gathering yourself. And it's all free so nothing better. Btw. I didn't realize cohete linked to a coursera course. There's usually free ones on there for most "coding" I think. You just have to do some digging. Additionally, they might ask you to pay for the certification or whatever. But who cares about that. You're definitely allowed to do the course for free on a lot of them and not even bother about the certification. It doesn't matter about the certification anyway because all anyone should care about is learning the material and acquiring the knowledge. Try these (should be free): https://www.coursera.org/course/rprog https://www.coursera.org/course/statistics The first one looks to be just programming based. Which is good if you already have the knowledge to do what you want to. I'd actually recommend the second one depending on what your prior level of understanding about math and statistics is. It'll teach you some of the basics on statistical inference. Seems like they do a thorough job, starting with some basic probability theory and then transitioning into statistics using the central limit theorem. And then you have all the bases covered with confidence intervals, hypothesis testing, and they go into some forms of regression. They also do chi-square and ANOVA which are important things to know. Seems like the only entry level thing not here is some stuff on time series. Overall seems really balanced and I think a good place to start by glancing at that syllabus. Anyway, best of luck.
Some of the quants at work use elements of R in their trading. They recommended to start here: http://www.r-bloggers.com/how-to-learn-r-2/
Damn. Really? I can't imagine why. I'd suspect that a combination of c++ and matlab or python would be much better in almost every scenario.
Not really sure. I'm not a quant. I just asked them if they use R and they recommended that link for beginners. I'm pretty sure they use the other programs as well though.
<iframe src="https://channel9.msdn.com/Events/useR-international-R-User-conference/useR2016/FiveThirtyEights-data-journalism-workflow-with-R/player" width="640" height="360" allowFullScreen frameBorder="0"></iframe>