Tips for beginners in R

“Tips for beginners in R” is the second article in the series. The first article can be read here – Why every business analyst needs to learn R?

In may last article, I mentioned R is the next big thing to learn in analytics. However R documentation is not so easy to pick up, and it is a more advanced language than other statistical computing platforms that analysts have been using. My advice on this is- Do not get overwhelmed by the flexible nature of R. R is supposed to have a steep learning curve, but some of the things can be learnt quite fast in R especially if you are a business analyst. Here are some basic tips, tricks and techniques to learn R as business analytics platform relatively quickly.

0) Download the software from the www.r-project.org website.

1) Read the basics of R from online manuals. You can read manuals on R at https://cran.r-project.org/manuals.html . The R Online Documentation is quite well organized.

2) Do not flood the R help email groups given at https://www.r-project.org/mail.html with questions unless you have searched the Internet for answers.

3) Command Line Programming can make a beginner uneasy when learning a new language. One of the reasons is trying to remember new syntax and commands. Fortunately r now has a great collection of Graphical User Interfaces of GUIs. Use some GUIs to move stratight to lcik and point analysis than be frustrated by learning about which command does what, and which package to learn.-

Choose Appropriate GUI or Tool

1. Deducer for Data Visualization – If data visualization and graphical analysis is what you are primarily looking for the Deducer GUI is one of the most appropriate.
2. Rattle for Data Mining – If you are going to build models, or do clustering analysis the rattle GUI has a wide array of algorithms that can be easily used for such purposes. It has 4 kinds of clustering methods, and different kinds of model building methods including regression, decision trees , ensemble methods and artificial neural networks.
3. R Commander for Statistical Analysis and Time Series (using e-pack plugin) – The R Commander Graphical User Interface is one of the most simple GUIs in R, it is one of the most widely used, and it is extensible by atleast a dozen plugins that basically help bring drop-down menu functionality for packages.
4. Red-R for Workflow based programming – This is a relatively newer project than others, but it is based on work flow programming.
5. Revolution Analytics Enterprise R for Business Analytics on large datasets (using RevoScaleR package) – Using the RevoScaleR package you can basically do analysis on larger  datasets.  Note that for ordinary R, you are limited to size of RAM for the size of the datasets on whom you can do analysis .You should also use Revolution Analytics Enterprise R if you need dedicated customer support.
6. R Studio for developers who need an IDE

4) Use CRAN Views for dedicated help on topics https://cran.r-project.org/web/views/ .

These include subjects from regression ,clustering , high performance computing, time series analysis and others.

5) Buy a good book for beginers in R. There are 115 books listed at https://www.r-project.org/doc/bib/R-books.html . I personally like “R for SAS and SPSS users”

6) Watch video tutorials on Youtube on specifics of R Programming. One such series of tutorials is at https://www.youtube.com/user/Tutorlol

7) Subscribe to blogs (almost 300 of them)  pertaining to R at https://www.r-bloggers.com/ . This website basically pulls in content from almost all the blogs pertaining to R and you can subscribe to it by email.

These tips were based on my decade long experience in analytics and more than 5 years of experience in learning R. Every day I learn something new in R and that is what keeps me excited and motivated to learn even more.