If there's something nerdy on the internet that involves graphs, plotting, numbers, and animated maps, you bet your bottom dollar I'm a fan. So when I came across this info graphic on this blog, I had to share with you. As you may or may not know, last year a squashed my 2013 goal of consuming 100 different kinds of beer. This year I'm bent on discovering whisky and doing tons of pull ups, you know, to keep things balanced. I wont bore you with the latter. Below is the article:
His analysis of the data led to the question: where did the source data come from in the first place? With some crowdsourced sleuthing, Christopher discovered the data comes from the first edition of the book Whisky Classified: Choosing Single Malts by Flavour by David Wishart. The story behind the data is quite interesting, and worth checking out if you're a whisky fan.
It turns out the data file Luba used comes from the first edition of the "Whisky Classified" book, and there were a few typos in the data to boot (for example, Bowmore had a Medicinal ranking of 1, but was actually a 2 in the book.) A commenter "Florin" at the Scotch and Ice Cream blog cleaned up the data and re-ran the analysis, and generated four slightly different clusters: peaty whiskies, ex-sherry whiskies, ex-bourbon / no peat whiskies, and whiskies with some ex-sherry blended in or with some peat. Extending the analysis to five clusters apparently succeeded in "separating the hard-core peated whiskies from the less-peated ones". - Here