Naked Data #91 — One night of data viz passion
Codebridge hosted the Cape Town Data Viz group for a brilliant talk by Alex Conway last night. For a bit of Naked Data variety, we’ll showcase some of the visualisations and tools that melted our minds last night. Regular programming will continue next week.
Conway describes data viz as “how you map your numbers into pixels”, and if you want to do that quickly and easily, check out Raw. Drop your csv data into a form field, and then select your chart from an excellent selection of D3 visualisations you won’t find in Excel. Conway describes Raw as great for quick prototyping before settling on the data and visualisation you want to tell your story.
“Good data visualisation is something made in D3, not made in Excel.” Excel didn’t get a whole lot of love last night, but D3 sure did. The data and graphing library built by Mike Bostock is, apparently, what serious data scientists use.
Dc.js merges D3 and another great library, Crossfilter. Together, they let you start building filters on your data, which in turn lets you explore your dataset and ask it real questions. Conway used an example from the Unearthed mining hackathon, and crossfiltering D3 became a bit of a theme for the evening.
If you’ve got a map file (GeoJSON, TopoJSON or a shape file) and don’t want to crank open QGIS to look at it, MapShaper lets you simply drag the file to the page, and it renders it, complete with data. Sweet!
This is one of Conway’s favourite visualisations, ever. So when Adi Eyal said he hated it, there was a bit of a tense moment last night. I’m going to weigh in here and say that the animation between all spending and discretionary spending makes this viz awesome.
How reliable are election polls? Rock ’n Poll lets you explore that question for yourself, with an excellent balance of narrative and exploration. It’s also uber-cool that you’re using your browser as a simulator.
Movie directors, movie stars, constellations. Gettit? Anyhow, a cool viz for movie buffs.
These compelling visualisations illuminate the story of the US drought, without losing the fact that it’s a complex issue.
Here’s that cross-filtering again, and a great example of D3’s performance with 7,637 items with 14 data points.
Moar cross-filtering! This time, building a tool that is truly useful, and not just pretty.
Visualise your data in 3D! (At least I think that’s what’s going on here.)
Using D3 for data analysis, and showing off some algorithm skills.
Bostock takes D3 through its paces, transforming between different visualisations for the same data. Much eye-candy.
Useful for heatmaps, or if you need to visualise a bee hive.
One of those D3 visualisations that Adi complains about — it exists to show how clever the author is. It’s still awesome.
This is going to be a hit for those budget breakdown visualisations.
Voronoi makes it much easier and intuitive to select points with mouseovers. Conway showed us this map, as well as this multi-line voronoi.
Using Twitter geolocation, and working out whether a user regularly tweets from a location, this viz finds where the locals hide, and where the tourists hang out.
Whenever I do a presso on data visualisation, I haul out Adrian’s dotmap. Seems like I’m not the only one ;)