I'll tell you a little secret.
See all of those blog posts below this one? The ones that were written with dates in the past?
I wrote those ages ago and didn't post them. They're kinda fake.
Anyhoo, now they're real and published. rpackages now has an actual blog for me to write this stuff in and for you to comment. Isn't that amazing?
Code snippets now have syntax highlighting.
Don't tell me that I don't love you.
Today is Australia Day, and that means that while everyone else is having barbecues and drinking beer at 10am on a Monday, I'm hacking away at this website. Wooh!
I've wanted to collect more data on what people actually want out of this website for a while, and today's update is another major step in that direction. With it, I add the 'Vote for new website features' sidebar. It's meant to be a super easy way for you to tell me what is interesting to you. The features with the most votes get implemented first. Easy!
I've listed a few features that are interesting to me, but I still want you to contact me with your requests.
This website has a wonderful search function. Except that I, er, removed the search box. It's back now. Enjoy!
My wife gave birth to our beautiful new son James today.
He's pretty cute.
We're now indexing packages from Bioconductor and R-Forge. Tons of good stuff in there. You have no idea.
I've added a very basic search function to the site. You can search over package names. This is great if you have a rough idea of what you're looking for.
It's also OpenSearch compliant, so if you're using a suitable browser (e.g. Chrome), it'll let you search directly from the address bar. This is super convenient. If I'm looking for documentation for
mixtools, I type R-P-A, Chrome autocompletes the rest, TAB puts us in the search box, and then I can type 'mixtools'. Easy peezy.
My name is Ian, and I'm the developer behind rpackages.
I started this site because I had two major frustrations:
- Reading documentation for R packages was more difficult than it needed to be. I was sick of trawling through PDF documents which didn't put the information in the order that I needed it.
- Often, I'd install a package and find that it didn't work, or often didn't even compile correctly.
I started pulling out the documentation and compiling it on my laptop so that it could be indexed locally, and realised that other people might be interested in the same thing. I also started some VMs to build the packages so that if nothing else, I had advance warning of problems before I invested a lot of time into a package.
I want R to be treated as seriously as any other commercial statistics package.
I've argued with companies many times who won't take R seriously because it "isn't validated" or "isn't tested". And that's a valid complaint! But if you use SPSS or SAS for only a few minutes, you'll realise that they are also as buggy as sin -- they just have the 'validation' tick.
This site is my contribution to getting R taken seriously. We test the packages before use. All of them. We write acceptance tests. And if necessary, we certify combinations of core R and the packages as 'validated' and use them for our work.
I have a long, long list of features for this website. I want you tell me your ideas, too! Send me an email, a tweet, or just leave a Disqus comment at the bottom of any page. I promise, I read every single one of them.
In the near term, I want to:
- Get all of the CRAN package documentation fully loaded, compiled and pretty on the web.
- Add a search function over the package names and the package documentation. That way, to look up something, you'll be able to just type "rpa<tab>" in the Chrome address bar, and your documentation will pop up.
- Start publishing test results for packages.