Computing has always been an essential ingredient of statistical
practice. While probability theory provides us with a mathematical
foundation for describing estimation methodologies
and studying statistical inference,
computing technologies act as a medium
through which analyses are actually realized. Our ability to
manipulate data and to audition new methodologies depends on and is
limited by our familiarity with computing technologies. To some
extent, even our notion of what constitutes "data" is a product of
our background in computing.
Through a series of projects, we will study tools for
"exploratory computing." We will emphasize programming and scripting
languages over point-and-click interfaces. We hope to instill a
problem solving ability so that you will learn languages on your own,
cull online documentation or tutorials, find books and manuals.
This year, we have been awarded an Amazon Web Services educational grant
and most of our computations will be performed on
their Elastic Compute Cloud service. This approach will provide us with
a degree of uniformity in setup and (more importantly) will allow us
to audition tools for large-scale data analysis.