Normal deviance: My data analytics New Year resolutions

I'm not big on New Year's resolutions, but it is helpful to periodically stop and reflect on whether there are better ways of doing things. When reflecting on how to improve my analytics work, there are a few things that spring to mind.

1. Be prepared

A woodsman was once asked, "What would you do if you had just five minutes to chop down a tree?" He answered, "I would spend the first two and a half minutes sharpening my axe."

Ultimately much of analytics is about efficiency - improving the speed and scale of a process. One of the best ways to create efficiency is to step back and consider the process that could be abstracted or generalised. Conversely some of our analyses can be limited by the tools we have; and feature development can be significant in improving our work.

Whether it's setting up a new modelling technique, or systemising a data flow, or developing some targeted diagnostics, there are lots of ways to make tomorrow a little bit easier, and ensure the axe is sharp before it gets put to use.

2. Learn some new packages

The growth in packages; pieces of code for solving particular problems, has been near exponential. On CRAN, the network that manages R, there are now over 13,000 packages plus many others that live less formally on GitHub accounts. Moreover, the majority of these are less than five years old, which means keeping up to speed with new developments are a challenge. But keeping an eye out for particularly useful packages is still important.

Fortunately, there are good ways to cheat on this one. In 2018, the Data Analytics Seminar and the analytics video competition were both useful in highlighting a couple of newer packages that are relevant and interesting.

3. See the world

There's a lot of good conferences out there. The Actuaries Institute will continue to run a wide range of events across a range of topics, including the Summit in June. There might even be a couple of conferences beyond the Actuaries Institute that would be worth a look. These are often listed in Analytics Newsletters, archived here . In any case, it never hurts to plan so that you can navigate the twin perils of too little exploration and conference overload.

4. Document my code better

It's always good to have a stretch goal.

Whatever your goals are, I wish you a successful 2019.