Dataviz of the Week: Periodic Tables
Not long ago I got an e-mail from a programmer/scientist by the name of Michael Dayah who was understandably proud of his interactive Periodic Table of Elements.
The PT, I should say, creates an extraordinary opportunity for datavisualists. The names and properties of elements constitute multiple interlinked datasets which can go as deep and wide as the creator wishes. Visuals and links can can vividly demonstrate properties, if/then scenarios and complex relationships. Popups and tricks with Flash and Java can give the tables a sheen of awesome cool.
I’ll mention just two PT projects today.
This is Dayah’s creation.
The interactive version of the grab above lets you mouse over each element and, say, adjust a slider to see what happens when each element is heated to a given temperature. Mouse over various other prompts [which, sorry, I don’t understand well enough to describe] to see little illustrations of, say, “orbitals.” In Wikipedia mode each element is linked to its Wikipedia entry.
This application passes my first two tests for great data visualizations: It makes things clearer than any static presentation of data can, and it is engaging enough to invite extended exploration and serendipitous learning. A bright kid in junior high and an advanced chemist could equally get lost in the thing.
I found another interactive PT that passes my third test for great dataviz — simple awesome beauty–yet fails the first two.
This Visual Sciences PT, a product of the Royal Society of Chemistry, is [in Flash mode] gorgeous to look at. It offers illustrations (or artful symbols) of how the element is used or at least appears, permitting visitors to understand (a first for me) how those elements correspond to what most of us call “reality”: Erbium is used in pink-colored ceramic glazes! Who knew? The black background makes it look as slick as an ebony Boxster just out of the detail shop.
The trouble is, go one click deep and — with the exception of a blistering-cool but frankly inane Quicktime animation which pans an element image to a woo-woo audio clip–you’re in a static HTML page or a PDF. Well-edited and -presented, but dead text. Behind its shimmering exterior this application delivers 2-D information. No relationships, if/thens or serindipity. An eye-candy catalog of chemistry.
And so it turns out the homelier dataviz — Michael Dayah’s home-rolled but deeply interlinked version — is the superior application.
I’m always looking for dataviz that meet all three criteria: (1) the ability to bring instant clarity to complex concepts and data; (2) an implicit invitation to play and learn deeply; (3) beauty so striking that you just want to spend time with it.
Anybody have nominations?
Tags: datavizYou can comment below, or link to this permanent URL from your own site.