1.2 About Me

I began using R in my second year of university, during an internship looking at publication bias correction methods. I was under the tutorship of a member of staff who helped me immensely, but I must confess that I have never taken an official course in R, online or in person. I like to think, however, that this is not always a bad thing. Learning from the bottom up and struggling along the way is a fantastic way to acquire knowledge and instills a very important lesson:

You're not going to know everything there is to know about R. Ever. But that's okay.

I'm now 4 years into my R career and I use R every day. With that in mind, I don't think there has ever been a day where I haven't referred to a tutorial, or Stack Overflow, or even just Googled the name of a function that I've used 1000 times before. There is a great repository of knowledge for R and it's one of the things I love most about the R community. So please never feel as though you're an impostor in a world of R gurus. In reality, everyone else is just as lost as you. But if you keep ticking along and never feel that learning something new in R isn't worth your time, you'll end up doing some great things.

And in a roundabout way, that is part of the reason I decided to develop these materials. I don't pretend to be the ultimate R programmer, because I still know what it's like to learn something from the start. And everyone has to start somewhere. So I hope that I can help impart some of the lessons that I've learnt over the 4 years to anyone who's looking to learn R in a way that won't leave you feeling lost.

The only final note I have before we start learning how to use R is another bit of advice:

Don't believe everything you read

Whilst this is probably a good thing to keep in mind for any type of training, I feel it's particularly relevant with R for two reasons. Firstly, when it comes to programming languages, lots of people have opinions. Some are true, most are not. Most things you read are a mix between fact and opinion, so take everything with a pinch of salt. For example, the developers of the ggplot2 package are fervently against arbitrary second axes and so support for them in ggplot2 is limited. I also share this view, but that doesn't mean that I'm right - read, learn, but question and make your own mind up.

Secondly, R and particuarly all of its packages are prone to change. For this reason, people may make statements relative to one version of R that aren't necessarily true in the future. Things have changed over the years, and so answers from a 10 year-old Stack Overflow question may not still be true when you come across them. A microcosmic version of this are some recent changes in the tidyr package. Historically, converting data from/to long and short formats was done using the spread() and gather() functions. However, in newer releases, these functions are deprecated in favour of pivot_wider() and pivot_longer(), which provide the same functionality but also some extra bits. The practical implication of this suggestion is don't always read one tutorial on a subject before you dive in.