In my posts to this blog (less frequent than I would like, hopefully more frequent in the future), I’ve often mentioned my R package regtools, which contains a number of functions useful for regression and classification. None of them duplicate what is available in the excellent packages on CRAN, so I will dare characterize regtools as innovative. 🙂 In spite of the smiley emoticon in that last sentence, I hope you feel the same way.
You can download the package from CRAN, and yes, Joe R., it includes a vignette. 🙂 It begins with an example of one of the main features of the package, the use of nonparametric regression estimation to assess the fit of parametric regression models. I think/hope you’ll find it interesting.
Mad (Data) Scientist 
Congratulations, Norm! I’m a loyal reader of this blog and I’m waiting for your book! 🙂
Thanks! I’d guess mid-2017 as a publication date.
You have four (FOUR!) simple pull requests on Github: https://github.com/matloff/regtools One is months old, two are a couple of days old, and I just sent another one. Such simple pull requests should have been merged 5 minutes after they were sent to you!
At some point I want to write a blog post about the psychology in programming, of which I think many programmers have not realized the importance. Merging simple/trivial pull requests is a fantastic way to encourage users to contribute to your project, and it is an important sign showing your attitude to social programming.
Anyway, I’m very excited to see your move to the world of Github, R Markdown, and package vignettes! There is another thing I was wishing you could adopt, but it is not mature yet, so I’ll tell you the secret a couple of months later 🙂
I don’t mean to sound defensive (though I will). But Yihui, I envy you for having such a single-task job, as opposed to the absurdly multitasked professional life that I lead. 😦 It’s a great life, one that I feel extremely fortunate to lead, but it’s tough to do every little thing. None of those pull requests involved fixes/changes/enhancements to code, just typos and links. In spite of all this, yes of course you are correct. I’ll try to stay on top of this in the future.
Yeah I admit my life is much simpler than yours. My working hours can be pretty much committed to writing code. There are no administrative tasks on my shoulders, and there is zero bureaucracy at RStudio, so I’m always highly motivated to do what I really love. To me, multitasking basically means different projects/packages. I found it very helpful to involve other people in my projects by motivating them through trivial contributions in the beginning (e.g. fixing typos). If I happen to be too busy to maintain certain packages, I can hand them over to other people who have made substantial contributions in the past. This is also important for me to focus on more important and interesting projects.