Keeping up to date with R news
I've now given my talk about "How to be a resilient R user" three times, at R-Ladies Strasbourg and R-Ladies Paris in person, and at R-Ladies San José via Google Hangouts. It was fun! I covered part of the content of that talk in a blog post about where to get R help. Today, it's time for a post full of my personal opinions! I'll cover the rest of the talk: why and how stay up to date with R news?
Why it's important to stay informed
Following the news is a big part of being a resilient R user, because it helps you solve the problems you didn't know you had (maybe you'll see a package that can replace a cumbersome script of yours!) and the problems you will have (one day, you'll need to make a map, and be thankful you already vaguely know
sf exists). Obviously, staying aware helps because you learn about tools and their applications. But it also helps because it makes you learn about people and organizations! It is highly valuable to get a sense of who's working on what and of where the developments on a topic happen, because when you consciously look for something later, it'll help you find your way more easily.
Where to get your R news fix
Here is a probably non exhaustive list of where to get R news. Pick the ones that fit your interests and workflow better, and do not hesitate to tell me of other information sources in the comments!
Infinite information flows: Twitter, GitHub and dev.to
Both Twitter and GitHub offer some sort of timeline, as does dev.to.
On Twitter, there's the timeline made of content (re-)tweeted (or liked, thanks to that new random aspect of Twitter...) by accounts you follow. You choose your own adventure by selecting which accounts to follow! There's also the #rstats feed, either recent tweets by popularity or all tweets by publication time. I like browsing the latter to see interesting questions, news or applications of packages, R-Ladies boasting, etc. Keys to a good experience are, in my opinion
- Muting the spam and accounts you do not like. In particular, keep the habit to mute the big data blabla companies that start polluting the feed, you'll see gems more easily.
- Not being a completist. I've been told it's stressful to see the flow never stops. Well, make peace with that, you won't read everything and you'll be fine.
- Practice! I know I've gotten more deliberate, and much faster at browsing the content. Interestingly it probably means that content producers need to be wise in their choice of emojis, illustrations etc. to catch people's attention, but that's not today's topic.
On GitHub, I only use the timeline fed by accounts I follow and repos I watch: repos created or forked, new releases, people following each other (kind of cute when you see long time R friends following each other!), repos starred. It can help discovering new code, and new people. By the way, adding an avatar and minimal information to your GitHub account is crucial in general, and useful for the few seconds of screen time you get on someone's timeline after being followed by someone they follow: if your account is empty, nothing is read, otherwise, people will see "Jane Doe started following Ada Lovelace, computer scientist" which is more informative than "Jane Doe started following alvlc".
Note that there's no ad on GitHub timelines... yet.
<script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
When reading my GitHub timeline and seeing what I think are people experimenting with #rstats usethis/devtools, I wonder whether the default description could be sold as ad space— Maëlle Salmon
😉pic.twitter.com/eUVkUqdZyX 🐟(@ma_salmon) 9 de març de 2018
- Thanks to rOpenSci's excellent newsletter maintained by Scott Chamberlain, I've learnt about dev.to. Scott wrote "A shout out to a probably lesser know place to share and gain R knowledge: If you haven’t heard of dev.to, it’s a sort of developer forum for sharing articles and discussing. They have an R tag https://dev.to/t/r and an rstats tag https://dev.to/t/rstats - the former has more traffic." I have not started reading it myself, but it's good to know!
This is a category I made up, where I'm putting these information sources that are less overwhelming than Twitter because already more organized and curated.
R mailing lists could be your thing, be it one of the field specific ones, or the more general R-announce.
R Weekly is a great newsletter organized in categories making it easier to browse it and select what picks your interest. The very first section, highlights, is made of a few links on which R Weekly team voted. You can subscribe by RSS feed, email, or follow the Twitter account. R Weekly content stems from RSS feeds R Weekly subscribes to, and pull requests. Note that if you want to know how to get your content on folks' radar, you should check out this post of mine about blog promotion.
rOpenSci biweekly newsletter is as written earlier excellent. It includes news about rOpenSci packages, their application in scientific articles and in blog posts, reminders of recent content on rOpenSci blog, and since this week related news (that's how I read about dev.to), as well as call for maintainers.
R-Bloggers aggregates content from R blogs, re-rendering the whole content on their website. It can be a way to discover interesting blogs.
You could create your own news aggregators made up of RSS (or JSON?) feeds of blogs and newsletters you enjoy. I don't have a routine for that, but have heard of Feedly and co.
R conferences and meetups
R conferences (satRdays, useR!, RStudio::conf, LatinR, ConectaR, etc.) and meetups (R-Ladies meetups, R User groups) will allow you to stay aware of new developments as well as to make R friends. How to know which ones are taking place?
R Weekly has a section about upcoming events.
There's a weekly post about events on RStudio community forum.
Jumping Rivers maintains a list of R conferences and local user groups.
Now, attending events is obviously fantastic, but you can also make the most of what's online!
Conference programs can give you a good insight into what's new, trendy, and interesting to catch up on. Especially topics you see in several programs.
Slidedecks are often made available and links aggregated on the conference website or by volunteers (here's a links collection for RStudio::conf 2019 maintained by Karl Broman).
At this point, I'd like to add a short side note about R-Ladies abstract peer-review system organized by Jennifer Thompson! The idea is to encourage R-Ladies to speak at R conferences by giving them feedback on their abstracts. R-Ladies, submit your abstract. R-Ladies and allies, volunteer as abstract reviewer.
Other sources of information
Maybe you're into podcasts? If so, check out e.g. Not So Standard Deviations, Credibly Curious and The R Podcast. The latest R Podcast episode is a crossover between the three since R Podcast's host Eric Nantz is joined by Credibly Curious co-host Nick Tierny and Not So Standard Deviations co-host Hilary Parker! Jumping Rivers maintains a list of data science podcasts.
I like this monthly post by RStudio Joseph Rickert, selecting 40 packages among all new CRAN releases, organized in sections which makes it easy to digest.
What to do with the information
Now that you know how to get a ton of info, what should you do with it?
Assess the information you get much like you would assess any piece of news, and maybe ask your R friends what they think. If hesitating to try out a new package, always perform a mini assessment to estimate whether it's worth your time and trust. Assessing a package is not the topic of this post, but have a look at e.g. criteria used in this rOpenSci unconf project, Jeff Leek's post about how he decides to trust an R package, rOpenSci packaging guidelines.
Store it somehow
Maybe you don't need any system because you have a good memory, maybe you do need one. I'm not an organization coach so will only throw a few links/ideas. Figuring out what works best for you is very important, it'll make this time investment reading news worth it.
You can search your GitHub starred repos so starring a repo is a bit like bookmarking it.
Lise Vaudor wrote a blog post in French about se souvenir de tout, including a pic of the cheatsheets near her desk.
Maybe needless to say, if you like something you read or use, go forth and post it in your networks, to help feed the information flows. :-)
In this post inspired by my recent R-Ladies talks, I summarized my views on why, where and how to stay aware of news about R. I had included the topic in my talks because I've been asked "But how do you know about this?!" and answered "Oh, I simply read about it". After a few similar conversations, I realized that maybe it'd be useful to share my personal experience. I hope you find your own optimal routine, discover awesome stuff... and remember about it when needed!