Efficient R Programming: A Practical Guide to Smarter Programming 1st Edition

Simply a great book, chock full of tips and techniques for improving one's work with R.

Having written my own book on code optimization, I have a little bit of bias on the right way to do things. I wasn't even a little bit disappointed by Colin Gillespie's book on R. Instead of diving in to code smells and best practices in coding, we roll back to the beginning. How did you set up your R service? How is the system configured? Getting the basics right before tackling coding is absolutely the correct way to get things done. Further, while there is a lot of time spent on coding correctly, memory management and all the rest, the focus, again and again, is on all the methods, not just coding, but data movement, chunking up the work, talking to the business. These are every bit as important to efficient coding as using the correct functions. He even goes a step further and gets into collaboration through code formatting and source control as an aspect of efficient coding.The book is extremely concise. We get an example or two and a short explanation of why we're doing things and then it's on the next topic. If you're just learning R, you'll need another resource or two. Though it may be a little short sometimes, it covers a lot of material in only 200 pages. I found myself going back to sections of the book several times to better understand the points being made and because there were so many points in a short space.Overall, I liked the book and the writing style. I learned a ton about better ways to write R code. I'm going to use the information in this book. That's the highest praise I can give it.

The content is reasonable, but the book is very short and wow is the editing poorly done! Typos and other errors on every couple of pages, and the layout of figure legends seems to to have been considered at all. The amateurish production undermines the authority of the book’s tone, and it comes across in the end as a self-published pamphlet that someone decided to quickly turn into a book. I’m very disappointed with O’Reilly’s lack of attention to the book’s production. The material, the author and the purchaser deserve better.

On page 40 you will learn about keyboard shortcuts in RStudio. On page 50 there are a few pages about vectorization. On p. 154 a byte is defined. Too much of the book is at this level. There are bits and pieces that I find useful (I didn't know about the feather file format, for example), but on the whole this is much more basic than the title would indicate.

Writing beautiful concise code is definitely an art, and one that can be difficult to master as so many of us end up pulling together code to perform analytics in a very quick and dirty manner. This reviewer has been guilty of not commenting code and writing redundant routines because it was easier than figuring out the error in a macro or loop. In this book, Gillespie covers the concept of efficiency both through the amount of computational time and the programming time. For large datasets with a "standard" compute, in particular, R can be a little feisty, so computational tips are always useful.From that perspective, this book assumes at least a basic familiarity with R, although it touches on many of the intro concepts. It's not a good place to start if the reader is wanting to *learn* R. The concepts in it are also applicable to other languages, so it doesn't have to be R specific, but the code snippets are all designed to be executed in R. It links to other resources to learn R (including a personal favorite, the R Inferno), which is a nice touch. There is some discussion on the differences in R set up between OS (Windows, Linux, MacOS, Ubuntu).The tips are sometimes interesting - e.g. vectorize data whenever possible. Eh, there are some data types where that's not possible and perhaps not indicated for the types of analysis that need to be performed. Are there other approaches that could work? Those are not to be found in this book.Overall, moderate beginners and some intermediate R programmers will find this useful. Self-taught R programmers will also find some nuggets. Worth a read.

Other parts don't.The examples using datasets sometimes are complete, sometimes the required text is written later. And some point to outdated linksThe optimization chapter fails because the source code cannot be retrieved

Probably not necessary unless you are running REALLY large data sets. On the other hand, programming elegantly is useful when you need to make changes later as it is easier to figure out what you did in the past. You do not need this book right away if you are just learning R, but if you decide to do massive data mining it could certainly be worth the effort to master. This book would be helpful in that case, it is well written and clear if you have a medium level of knowledge in using R and statistics.

When looking at this, I expected it to include more for folks beyond the basics, which this seemed to cater to. For example, as another reviewer noted, a "byte" is defined on pg. 154, which is pretty simplistic. There are other examples of that, too, and they happen with a bit too much frequency. Also, this relates to large data sets, so a lot of folks will simply take a few extra minutes and approach these subjects differently.That said, I gave it a 3.5 star score because this did have a few portions that were helpful and that I learned from. After running across one of these, I actually stopped and looked the book over again, and I found it a bit better were I looking at it as a beginner or as someone with a specific need set. There was also discussion on items that helped because of their approach, but this is pretty much specific to how you use R and what you want/need to learn.