Bit Of A False Start
Pretty typical.
In my current role as an experimental scientist I often find myself needing more knowledge, sometimes in new fields. My typical modus operandi - and I believe this is common - is to go out and find a couple of resources (online, in print, colleagues) and tap them for information. Using a couple of different sources gives a rudimentary "error bar" on the information. Agreement on a topic? Maybe reliable. Vastly differing opinions? May be worth finding additional sources.
When it came to learning new skills for data science, I found online polls, books, introductory lessons, and devoured them. Just dive in, right? As I began to read through DAwOST, I realized that I'd most strongly benefit from being able to recreate the various graphs and data representations that are presented in the text. With the simple tools I had so far (e.g. Codecademy Python lessons), I couldn't do anything like that. This was a great eye-opener and suggested that I take the time to build some foundational skills and fundamentals before getting too far ahead of myself.
So, picking a language (perhaps Python, for no better reason than its popularity in data science) and learn how to a) use run it locally, on my own machine b) handle input files, i.e. arbitrary data and c) make some basic graphs have moved to near the top of my to-do list. And in the "general skills" category, I've also enrolled in another Coursera course on basic statistics. I think it will be great to see how these concepts are traditionally taught (hopefully in some sort of logical order), but it also has a focus on programming in R, another language with which I want to be familiar. (Not enough spare time for that; my thesis deadline is knocking.)
Then, I look forward to punching into the meat of DAwOST and also beginning to play with data on my own.