Best websites for R Studio / statistics help?
December 12, 2017 2:07 PM   Subscribe

I am brute-forcing my way through learning R for a big project I am working on. I am looking for your recommendations for the best "R" advice websites I can troll (and possibly post in) for help. I'm familiar with Stack Overflow, but it seems intimidating for a n00bie like me. In your experience, what are good places to ask statistics (particularly around general linear modeling, nested effects, and regression) and R questions?
posted by absquatulate to Computers & Internet (13 answers total) 27 users marked this as a favorite
Best answer: R is difficult to specifically search for in search engines sometimes if you have a particular question, so I recommend for searching.
posted by foxfirefey at 2:09 PM on December 12, 2017 [6 favorites]

The RStudio community site is good for asking R questions, and the people there tend to be friendly:
posted by thrungva at 2:45 PM on December 12, 2017 [2 favorites]

The term "rlang" can be more effective in searches.
posted by sammyo at 5:38 PM on December 12, 2017 [1 favorite]

Best answer: Stack Exchange can be intimidating, but if you make it look like you did some work before asking the question, they hopefully won't be snide pedants. Like post a snippet of code you're trying to use, and explain what you'd like to do. I tend to get good responses.

If you aren't using Hadley Wickham's packages collectively known as the tidyverse, YOU MUST. They make R so much better.
posted by vogon_poet at 8:11 PM on December 12, 2017 [2 favorites]

Best answer: Also, check out:

Which offers some more concrete answers to your question.
posted by vogon_poet at 8:14 PM on December 12, 2017 [1 favorite]

Quick-R is a good resource for sample code for data manipulation and statistics, though there isn't an associated message board.
posted by esoterrica at 8:27 PM on December 12, 2017 [1 favorite]

What site was helpful varied a lot for me even with closely related questions - I just googled a lot. It's all about finding the person who had your exact problem recently enough that the code transfers.

RStudio has a really helpful cheatsheet to use in conjunction with dplyr and tidyr (of the above-mentioned tidyverse), if you have data munging to do. (I also like ggplot2, but I got through that with googling.)

I also referred to Discovering Statistics Using R in hardcopy a fair amount, but it's expensive, the authorial voice is really annoying to a lot of folks, and I probably could have done without.
posted by momus_window at 9:12 PM on December 12, 2017 [2 favorites]

I mostly find myself on StackExchange with R questions. The authors of some of the packages I use most often hang out there. There's also the option of asking your question on Twitter with the #rstats hashtag (many package authors and RStudio people are also active there).

Also, nthing the Tidyverse love.
posted by pemberkins at 10:35 PM on December 12, 2017 [1 favorite]

Response by poster: Thank you everyone; I will take a look (and muster up some courage to post at Stack Overflow... Just need to know why when I add a nested random effect to my linear model, it says my eigenvalue ratio is large and can't identify the model... *sigh*)
posted by absquatulate at 6:14 PM on December 13, 2017 [1 favorite]

Best answer: I find R-bloggers does a good job of providing example code and clear explanations.

I am also not too proud to google the exact error message, maybe with another key word about the method. Poking around, I found this. Good luck!
posted by MrBobinski at 5:26 PM on December 14, 2017 [1 favorite]

Response by poster: Oh. My. Heavens. Thank you MrBobinski! I've been searching fruitlessly for this... guess my search terms were too vague.
posted by absquatulate at 12:15 PM on December 18, 2017 [1 favorite]

As far as references, I’ve found the book Discovering Statistics Using R (mentioned above, though mine is paperback) to be helpful. It covers some fairly advanced topics. Also, I have found Datacamp’s R tutorials useful. Most are not free, but the introductory month membership fee can be pretty cheap.
posted by reren at 8:16 PM on December 18, 2017

« Older Using piping with *nix find without getting mired...   |   Easy VHS -> MacOS capture? Newer »
This thread is closed to new comments.