Exploratory data visualization

At the outset of any data science project, there is likely to be a period of iterative data exploration when you gain insight into the data. Throughout this book, Incanter has been our primary visualization tool. Although it includes a large number of charts there will be occasions when it won't contain the ideal chart for the data you seek to represent.

Note

Other Clojure libraries are stepping in to offer exploratory data visualization capabilities. For examples, see clojurewerkz/envision https://github.com/clojurewerkz/envision and Karsten Schmidt's thi-ng/geom at https://github.com/thi-ng/geom/tree/master/geom-viz.

For example, back in Chapter 1, Statistics, we used a scatter plot with alpha transparency to visualize the voter turnout proportion against the proportion of votes for the winner. This wasn't an ideal chart, because we were primarily interested in the density of points in a particular area. Alpha transparency helped reveal the structure...