Welcome to Spring!!

I’d like to try to bring a little more structure back to R Club this term, with the hope that more structure will translate to more usefulness/learning/joy. So here’s an idea: Everybody (especially everybody taking R Club for credit) comes up with one or more topics for discussion, and we’ll spend some time on these each meeting. Individuals would sign up for particular days, and be “ready” to lead a little talk about whatever topic they’ve chosen on that day. We can figure out schedule stuff in week 1.

Good topics might be questions for the group (e.g. what’s a good way to get descriptive statistics for a dataset?), tips or tricks (e.g. check out this great tutorial I found on the lme4 package!), or short lectures (e.g. I’ve been playing around with the caret package – let me show you what I’ve figured out). When thinking about what topics might be good, remember that there is HUGE variability in the amount of R expertise at R Club. Roughly half of our attendees are brand new to R with zero experience, and the other half range from folks who know some basics (maybe from attending R Club in previous terms) but not much more, to folks who have serious expertise in a focused area but lack broad knowledge (e.g. they use R for analyses in their lab, but they only really know how to do the one thing their lab always does), all the way to folks who are veritable R czars with fluency and power the rest of us can only dream of. So if there’s a topic that you’re interested in, don’t fret about if it might be “too easy” or “too hard” for the R Club audience – it will almost certainly be both. 😉 If you think it sounds interesting or useful, chances are someone else will, too, even if it’s super basic material.

I’m thinking these presentations would be maybe 5 minutes long, and we’d probably do a few of them at each meeting. This would provide a concrete excuse for us each to think about what we want to learn in R this term (nothing is more motivating than a looming presentation, even an informal one), but it should still leave us with plenty of flexible time during R Club for us to work individually. And, most importantly, it facilitates us learning from each other, which is pretty much the whole point.

Being “ready” to talk about your topic will probably involve a little background work, like reading R documentation, looking for resources online, etc. If your topic is a question you want the group’s feedback on, take a little time to define the problem well, come up with relevant examples, etc. Depending on your topic, you might also want to have a short demonstration prepared. For example, if you want to show us a function or package, you will probably want to have a dataset ready to play with, so you can pull it up on the screen and show us how it works. Or if you have a question about how to run a particular type of analysis, it would be great for you to have a dataset ready for the group to try stuff on. If you find useful resources during background reading, or if you have example code to share, I encourage you to write up a short post for the blog, so we can all access the relevant goodies.

There are also very practical benefits to adopting a topic-oriented structure for R Club: If you take some notes during other people’s presentations and write down your own thoughts while you’re preparing for your topic(s), you’ll be basically 100% done with your “what did I learn?” write-up for Sanjay by the end of term. Also, since R Club meetings will still allow a fair amount of free-work time, you could theoretically do all of your preparation work for your topic during the R Club meeting before you present. Bam! Efficiency.

Sound like a good idea?

To help get the inspiration juices flowing, here are some topics I think would be fun and interesting (feel free to use one or more of these, if you like, or come up with something else that would be fun and interesting for you!):

• Descriptive statistics in R
• for loops
• Introductions to a variety of GLM favorites in R: t-tests, ANOVA, ANCOVA, multiple regression, etc.
• Introduction to one or more of the apply functions (nice tutorial here)
• SEM in R (the Lavaan package maybe? I’ve never used it, but it looks good. Potentially useful tutorial.)
• ggplot2! Such pretty. Way fun.
• Machine learning in R, probably focusing on the caret package, possibly assigning reading from this book which is available in pdf form for free (yay!) from the UO library
• HLM in R (lme4 time! woop woop!)
• Manipulating dataframes (handy tools include reshape and plyr)
• Data management, metadata, etc. in R (Jacob, what’s that awesome package?)
• Shiny! or maybe Plotly!
• ???