Sparklines

Sparklines

Bill mentioned that Edward Tufte has posted a chapter from his new book, Information Design about how good information design can distill multiple parameters into immediately comprehensible and intuitive information (sounds like his other two books).

<blockquote

“Also check-out Page 6, where he displays the win/loss record for an entire season for the Baltimore Orioles (162 games) inline with the text. Also notice how he uses just one additional color (red) to highlight unique data. This is great stuff.”
—Bill Tani


The basic idea appears to be sparklines which are small, high-resulution graphs which can be easily scanned in much the same way our eyes read lowercased words faster than UPPERCASE ones. They are placed in the context of text almost like words in a sentence. I saw these all the time in neuroscience papers, except for the fact that they weren’t placed in a sentence.

This seems a good approach for any time you need to compare multiple time series of different quantities. I really like the discussion of placing color queues and how to quantify the aspect ratio.

(Check out the comments at the end.)

There is even a library for PHP. This will be great fodder for co-lapse.

Leave a Reply

Your email address will not be published. Required fields are marked *