Hello everyone, how many of you use R instead of SPSS and Matlab (and SAS etc)? I am ok with SPSS, not good at all with Matlab. Since I heard that R is very powerful and trendy, I want to learn to use R. Any advice on websites, blogs and books? My institution doesn't have any courses on R, so I'll need to learn it by myself. Thanks!

it depends what field you're in and what you need to do with it. I have an office mate who only uses Matlab and claims its going to take over SPSS - which I doubt. I mainly use SPSS (which now does a lot of the stuff R was needed for e.g. bootstrapping) and I also use AMOS and Mplus for structural equation modelling.

But if you're only going to need to do a t-test or something then you might as well stick to SPSS.

Hi Sweetleafy
I used R for my project as the stats I needed to do just weren't possible/reliable with SPSS, which is the only other program I would use really. I tried teaching it to myself from scratch at the start and found it almost impossible, but that could be just me. I'm not naturally a computer programme-type person, and found learning all the commands a bit too much. Eventually I found a friend who had gone on a course and was able to give me basic code I needed for doing my stuff, which I then modified and made suitable for my stats exactly. It was a bit easier this way, as I was able to work through examples until I figured out what everything meant and I was able to see how to change things to suit my own needs, but it was still a bit of a slog. If possible you should go to a training course, there are courses run throughout the country although they are quite expensive - perhaps you can blag some funding for it by promising to run some tutorials to your department when you return?
Overall I'd say if you are learning it by yourself, you'll need heaps of time and patience, work through loads of examples, and use the R website links for basic code templates for various tests. Good luck!

Heya. I use R, after a few years tinkering with SPSS/SAS. I find R very powerful, and it can do anything that any other stats package can, which is really convenient once you know which code to use. It is a steep learning curve to begin with, but once you get the hang of how the code and output work it's fine. I hadn't written code (or had much stats experience) before I started to learn R, so it is definitely possible to teach yourself from scratch! I'd recommend The R Book (Crawley) for a step by step introduction to stats in R. If you are in life sciences you will find the tutorials and worksheets at this page really useful - http://users.monash.edu.au/~murray/stats/downloads.html. This course is how I learnt R. Good luck!

I found the transition from Matlab to R was no problem.

Everything is difficult before it becomes easy.

I actually taught myself R by trying to reproduce various fancy plots from ggplot2. It was a fun way to learn as it's visual and rewards you with very nice looking figures.

======= Date Modified 15 Feb 2011 09:50:38 =======
============= Edited by a Moderator =============
-inappropriate content-