I [Heart] Charts and Graphs

January 25, 2007 10:43 PM

Data analysis, brought to you by Big Blue, is following a trend. Data has never been more social. Geeks and statistics groupies used to be isolated, but the internet is changing that. Ever pine for a pile of Excel spreadsheets? Have you tried running an ANOVA on a year's worth of traffic data? You're not alone. New sites add sociability to cold hard facts; take a look at the "YouTube for data" or IBM's Many Eyes. Both sites induce squeals of delight from anyone who's ever felt Tuftian. What's next? One word: infornography. Please, keep your Standard Deviation jokes to yourself.
posted by Monochrome (16 comments total)

It's my first post in the blue after 1.5 years of lurking.
posted by Monochrome at 10:46 PM on January 25, 2007

Interesting.
I particularly like network diagrams with nodes you can drag around, with weighted distances and collapsable nodes.
posted by jouke at 11:10 PM on January 25, 2007

I am doing network diagrams and STATA work right now, and I already did an analysis of MetaFilter awhile ago... does this make me a statistics groupie? or can I just stay a normal geek if I don't understand the underlying theory?
posted by blahblahblah at 11:47 PM on January 25, 2007

cor Neat Post, c

| Neat Post
-------------+------------------
Post | 1.0000
Neat | 0.8452 1.0000
posted by faux ami at 12:05 AM on January 26, 2007

I must have played with IBM's Many Eyes for an hour or two the other day – the first and only Java applet of quality I have ever beheld. VNC clients have been the only other java applets I've used for longer than the few seconds it takes to close the tab in a panic – and even those will be replaced with Flash soon enough.
posted by blasdelf at 1:22 AM on January 26, 2007

Very timely. I have been working on a massive data visualization project for the past three weeks, and have become very familiar with Swivel and Many Eyes over just the past few days. Good post. This is probably more familiar to everyone, but I've often found clowd tagging to be a somewhat useful tool when analyzing writing or speeches.
posted by psmealey at 4:13 AM on January 26, 2007

faux amy, you can't get a correlation from an n of 1!
posted by naturesgreatestmiracle at 4:43 AM on January 26, 2007

These are great. Many Eyes would be much cooler if you could merge datasets with some degree of automation, though. That would also enhance the collaborativeness of the endeavor. For example, there's an interesting graph of coffee imports by country on the site now. It would be interesting to see that per capita, and not surprisingly there's a population dataset on the site, too. But in order to combine them you have to download both datasets to your computer and Excel them together.
posted by grobstein at 5:30 AM on January 26, 2007 [1 favorite]

This might be a good thread to bring up NetVis: If, for some reason, you like analyzing your online relationships, this tool offers "Dynamic Visualization of Social Networks." Haven't used it, but it sounds like the perfect embodiment of, as monochrome says, "infornography." Or maybe just an interesting new way to play with data and learn a thing or two.
posted by Milkman Dan at 6:33 AM on January 26, 2007 [1 favorite]

clown tagging
posted by mendel at 6:34 AM on January 26, 2007

Woo, thanks for the post Monochrome. Many Eyes is in Alpha now -- we're mostly interested in getting feedback from people on new tools that they might want to use (e.g. grobstein's comment above), trying out different community features that will help people have discussions around data, and of course adding new and cool visualizations based on what our users ask for.

For the time being, though, I'm just sitting back and enjoying all of the cool and weird data that people are looking at. McDonald's as a big lipid ball, Co-occurrences of names in the New Testament, and Presidential campaign receipts and disbursements are some of my favorites so far. :)
posted by xthlc at 6:37 AM on January 26, 2007 [2 favorites]

Nice post!
posted by nj_subgenius at 7:13 AM on January 26, 2007

Bounce bounce bounce. Happy tkolar!

Thanks, Monochrome.
posted by tkolar at 8:04 AM on January 26, 2007 [1 favorite]

I wanted to go with 'Get Your Curious On', but the bloggers ran with the 'YouTube for Data' tagline. Fine, we'll take it :~)

Monochrome, thanks for including Swivel in your post. When we launched Swivel in December 2006 a ton folks said, 'Data = Fun? No way.' After seeing Swivel and now Many Eyes not only can we say, 'Data = Fun.' We can also say, 'IBM = Fun.' This is exciting stuff.

We hope a whole of lot of hidden value gets unlocked from the world's data because of sites like Swivel and IBM's Many Eyes.

Go data!
posted by brianmulloy at 9:38 AM on January 26, 2007

I like the pile of Excel Spreadsheets:
'Have you ever made a spreadsheet for your personal life?'
...well, duhh...
posted by MtDewd at 10:13 AM on January 26, 2007

grobstein, what you described about automatically combining data is a big part of the value of Swivel. mashing up data sets together.

for example this graph of wine consumption versus violent crimes is a mashup of two data sets uploaded by two different people and combined by a third person:

it's the beauty of many people working together to discover insights.
posted by brianmulloy at 3:51 PM on January 26, 2007