I'm learning some basic statistics with applications by following along with the old MIT 18.443 Statistics for Applications course on OpenCourseWare. Note that this is the Fall 2006 version of the course, which I picked because I liked the book. The newest version of the course (18.650 -- Fall 2016) has no textbook at all. But so far I like having a textbook.
The course was taught using matlab, but I'm using numpy/scipy/matplotlib for Python instead. The data files are in .mat format on the MIT OCW website.
I'm not actually sure what individual dependencies are needed for these python scripts, but I installed all the scientific/numerical stuff I needed at once by installing Anaconda.
I've "completed" (with lots of cheating) the explicit programming assignments in the homework. There are other book problems that may require some programming, so I may contribute more to this repo in the future.