From William Couch, Kristen Novak, Michelle Price and Joshua Hatch of USA Today, this tracker tool lets you compare ratings of past current and past presidents according to Gallup polls.
The Organization for Economic Co-operation and Development (OECD) makes a lot of world indicators available (e.g. world population and birth rate). Much of it goes unnoticed, because most people just see a bunch of numbers. However, the Factbook eXplorer from the OECD, in collaboration with the National Center for Visual Analytics, is a visualization tool that helps you see and explore the data.
Those who have seen Hans Rosling's Gapminder presentations - and I imagine most of us have - will recognize the style with a play button and a motion graph in sync with parallel coordinates and a map. Choose an indicator, or several of them, press play, and watch the visualization move through time.
Also, if you've got your own data, you can load that too, which is certainly a nice touch.
Say what you want about Michael Jackson, but there's no denying the great effect he had on the music world. In honor of the pop king's passing, practically half of The New York Times graphics department stayed up late last night building this graphic. It takes a look at his majesty's Billboard rankings over his career compared to other popular music artists.
Decade after decade Jackson produced numerous hit albums. Click through time to see the mountains of each. Timeless.
To the man, to the legend, who no one will ever be able to replace:
A large majority of us who have websites use Google Analytics as our traffic monitor, and why not? It's free, it works, and it provides loads of data on traffic, referrals, and our content. We can then make decisions based on that data, but the trouble is there's a fair amount of clicking before we get to the good stuff. Enter Dshbrd by DabbleDB. Yes, that's dashboard with no vowels.
The DabbleDB folks know data, and Dshbrd is no exception. Using data from your Google Analytics account, Dshbrd analyzes and finds the points of interest - and then shows them to you in a clear and concise way. I've grown incredibly tired of overused sparklines, but Dshbrd uses them well to show traffic trends alongside a vertical stacked area chart. The two are linked such that when you scroll over an event (e.g. rise in referrals from Digg), the area on the stacked chart highlights and vice versa.
View traffic from site referrals, search engines, and direct links or content popularity, etc. Basically, you can examine all of your analytics data in Dshbrd that you can in Google Analytics but in this new view. It might take a second to get used to time on the vertical axis, but once you get over that, this alternative interface is quite intuitive and more importantly, very useful.
Now if only DabbleDB would provide a reliable API I would be very happy.
Ultimately, I'm guessing DabbleDB would want to turn Dshbrd into a fee-based service if it gained enough traction. I personally wouldn't pay for it since I really don't need that much outside the usual Google view, but I could see how Dshbrd could be useful to others. What do you think? Would you pay for this sort of premium view into your Google Analytics data?
Axiis, an open source data visualization framework in Flex, was released a few days ago under an MIT license. I haven't done much in Flex, but from what I hear, it's relatively easy to pick up. You get a lot of bang out of a few lines of code. Axiis makes things even easier, and provides visualization outside the built in Flex graph packages. Continue Reading
Roland LÃ¶ÃŸlein, a media student at University of Applied Sciences in Augsburg, presents meteorological time series data in 3-D in a class project called Synoptic. Rotate and zoom in and out on the different time lines, select different metrics, and compare against the corresponding time series on the bottom. After a few minutes of playing with it, I'm still trying to decide whether or not it's useful, but I think it's more of an experience than it is an analytical tool. It's almost like exploring a map, but instead of rolling hills, you get dips and peaks in a chart. Interaction is smooth and the visualization scores well in aesthetics.
While on the topic of job losses, USA Today provides a look into job forecasts from Moody's Economy.com. While the new forecast shows U.S. employment growing in 2010-2012, the outlook for different sectors and states varies quite a bit. Take a look at different job sectors via bar chart and map and then filter down by state.
[Thanks, Juan & Ron]
There are so many ways that you can cut a dataset whether it be big or small. Cut it by time, different chunks of time, categories, etc., and you just might get a different story out of your graph. Over on Barry Ritholtz' blog, The Big Picture, debate over the extent of job losses and this recession led to these four depictions of, well, job losses and recessions.
GOOD Magazine's most recent infographic (above and below) on consumer spending got me to thinking about all the other approaches I've seen on the same topic. The number of ways to attack a dataset never ceases to amaze me, so I dug a little. Yeah, there are a bunch - but here are some of the good ones. Got some more? Leave a link in the comments.
I ran a contest last week to improve a graph from Swivel that showed immigration to the United States. FlowingData readers sent in lots of different approaches (that took me forever to get organized for this post), and I still stand by my statement that there's always more than one way to skin a dataset.
This is a guest post by Simit Patel of InformedTrades, which offers free advice on trading stocks.
While many investors use economic and fundamental factors to identify investment opportunities -- i.e. whether a company has good management and is in a growth industry, or how it will be affected by macroeconomic conditions -- ultimately the price of an asset comes down to two things: supply and demand. The demand for buying vs the demand of selling. By visualizing the movement of price assets, we can gain an understanding of the psychology of the market as a whole, and thus what direction the price will go.
In Google Flu Trends, Google uses related searches to predict flu activity in your area "up to two weeks faster than traditional flu surveillance systems." The above graph shows query-based flu estimates compared against flu data from the U.S. Centers for Disease Control.
We have found a close relationship between how many people search for flu-related topics and how many people actually have flu symptoms. Of course, not every person who searches for "flu" is actually sick, but a pattern emerges when all the flu-related search queries from each state and region are added together. We compared our query counts with data from a surveillance system managed by the U.S. Centers for Disease Control and Prevention (CDC) and discovered that some search queries tend to be popular exactly when flu season is happening. By counting how often we see these search queries, we can estimate how much flu is circulating in various regions of the United States.
When Google first launched their visualization API, you could only use data that was in Google spreadsheets, which was pretty limiting. Yesterday, Google opened this up, and you can now hook in data from wherever you want. What does that mean? It means that developers now have access to all the visualization API offerings like before, but it's now a lot easier to hook visualization into data applications.
It also means we're about to see a boom in web applications that look very Googley. Motion charts (above) are going to spread like wildfire and ugly gauges will grace us with their presence. It'll be similar to the Google Maps craze, but not quite as rampant. In a couple months from now, I will have a long list of online places that use the Google visualization API. It's going to be interesting where online visualization goes from here.
Going back to my original question, to what extent do you think the now-open Google Visualization API will affect visualization on the Web?
I just noticed that when you click on "show details" in Google Reader, you get a graph of how frequently posts come from that feed and how often you read those posts. It used to only show subscriber count (via Google Reader) and when the feed was last updated. It's one of those things where it's like "so... what" and it won't influence any of the decisions I make in life in any way, but hopefully when all of you "show details" for FlowingData all the red and blue bars are aligned :).
P.S. Greetings from Chicago. It is much too early in the morning.
This is a guest post from Michael Drumheller, Dirk Karis, Raif Majeed and Robert Morton of Tableau Software. They use Tableau to explore the relationship between polls and predictive markets.
Predictive markets such as Intrade and the Iowa Electronic Markets have attracted more attention this year than in past Presidential elections. Some political observers such as ElectoralMap.net look to these markets as indicators of who's winning or losing.
Memeorandum shows up-to-date posts from leading political bloggers, and it is well-known that political bloggers are often very partisan. It's not always obvious to new readers though which side of the line a blogger sits on. You certainly can't always tell just from a headline on Memeorandum. So Andy Baio, with the help of del.icio.us founder, Joshua Schachter, created a Greasemonkey script (and Firefox plugin) to do just that. Simply install the script and browse popular political articles by their bias.
With the help of del.icio.us founder Joshua Schachter, we used a recommendation algorithm to score every blog on Memeorandum based on their linking activity in the last three months. Then I wrote a Greasemonkey script to pull that information out of Google Spreadsheets, and colorize Memeorandum on-the-fly. Left-leaning blogs are blue and right-leaning blogs are red, with darker colors representing strong biases.
Just a quick glance at Memeorandum with the plugin installed shows the magic works.
Of course this isn't just magic. It's not human-powered. It's a data-driven algorithm. It's statistics. The data are the articles that the Memeorandum-listed blogs link to, so just imagine a giant matrix with number of links. They then use singular value decomposition (SVD) to reduce that matrix to one dimension which they use to estimate where on the political spectrum any given blog on Memeorandum sits.
All you statistics readers (and maybe some of the computer scientists) should be familiar with SVD. I learned about it and played with it quite a bit during my first year in graduate school. Anyways, it's cool to see statistics at work and how it can be useful in visualization. A lot of the time visualization projects are about getting all the data on the screen, but with a little bit of know-how (or help from someone who has it) you can produce projects that let the computer do a lot of the pattern-finding work and don't make the user work so hard.
By the way, Andy's blog Waxy has become one of my favorite blogs as of late, so if political bias isn't your thing, I'd still encourage you to go check it out.
A while back I asked what you wanted to see more of on FlowingData. Thanks to the 447 of you who responded.
I was actually kind of surprised that there were so many votes for statistical visualization. I thought there would be more of a balance between design, art viz, and stat viz. I was, however, happy to see that the second most voted-on choice was "All of the Above." I must be doing something right! So by popular demand, here's some statistical visualization.
Since the above pie chart is making some of you cringe in agony (although I can't imagine why), let's take a look a few alternatives for the pie chart using the same poll results.
How about a horizontal bar chart? The results are sorted and you can easily see the difference in voting counts.
The above bar chart is missing a little something though. It doesn't explicitly show that each bar is really a part of a whole - in this case, all the people who voted. How about a stacked bar chart then? It shows the groupings and is a little easier to read than the pie chart in the sense that it's linear differences as opposed to radial.
Let's not forget our friends the bubbles. Carrying the same "problems" as a pie chart, the bubbles on the left are essentially a table with some flavor.
Personally, I still like the pie. Which one do you think is best? Or is there something else that might have been better than the above? How about a mosaic plot? Donut graph? A plain table?
Gas prices have been pretty crazy lately. I'm not used to paying over $45 for a tank of gas in my fuel-efficient Honda Civic. I mean, come on, what the heck?
So naturally, we want to know, "What do the data look like for gasoline prices?" The Energy Information Administration has this data available for download. They have historic gas prices for certain states (not all, unfortunately) as well as for U.S. regions. Check out the animation showing the rise and fall... and rise.. and fall and rise of U.S. gas prices from 1993 up until now. Things started going crazy in 2006.
Are you ready for another deconstruct/reconstruct exercise? I just posted a time series plot in the FlowingData forums that shows suicide rates and unemployment rates in Japan. Here are questions worth considering:
At a glance, the graph almost looks fine, but on a slightly deeper than superficial look, there are some clear problems.