Category: Data Analysis

Mapping Why I Will Be Late to Work for the Next Year

Posted by – June 5, 2016

After decades of neglect, the Washington DC Metro system will begin a year-long surge in track maintenance on Monday, starting with two weeks of work on the Orange Line that will reduce service on my daily commute from Vienna by 70 percent:

This surge of repairs to the nation’s second-busiest mass transport system is in response to a growing series of service delays, track failures, and electrical fires culminating in last year’s death of a commuter trapped in a tunnel on a smoke-filled train. General Manager Paul J. Wiedefeld ordered the maintenance surge along with changes to the system’s budget, culture, and management in an effort to improve safety and reliability.

As well as making commuting a challenge for the 6 million residents of the Washington DC region, it will be something of a social experiment. Commuters may choose to work from home on their companies’ or agencies’ VPNs or email, but mine gets overloaded when Federal offices are closed on snow days and it’s impossible to work on classified systems outside of the office; how will that change the pace and content of work, office culture, and family life? Metro users may get up earlier and arrive home later, or shift to other modes of transport; will this result in massive traffic jams, increased air pollution, more accidents, and a boom for bicycle shops?

The Washington Post’s SafeTrack web site provides further analysis and graphics, and the Washington Metro Transit Authority’s website is a good source for up-to the minute reports. I’d also recommend the outstanding free app DC Metro and Bus by DixonMobility (iOS and Android) to tell you when you’ll get to squeeze onto that next train…


Record Warmth in 2015

Posted by – January 31, 2016

AGU reports that NASA and NOAA independently assessed that the planet’s average surface temperature jumped to a new high in 2015, except for a cool spot in the North Atlantic. Based on about 6300 stations, ships and buoys, both agencies found the average global temperature had increased compared to 2014 – which had also been a record year. NASA found the average was 0.87°C higher than the average temperature over 1951 – 1980, where NOAA found a 0.90°C increase compared to the average over 1901 – 2000. The below image, courtesy of the Scientific Visualization Studio at Goddard Space Flight Center, shows the 2015 change from the longterm average temperature, with hottest areas on Earth in red and the coldest in blue.
Global average temperature 2015
Compliments to Cody Sullivan, AGU Intern, for the excellent article in EOS this past week!


Have a Data-Driven Christmas

Posted by – December 18, 2015

In a recent interview with Santa, IBM’s Graeme Noseworthy discussed the North Pole’s use of Big Data and analytics to assess who is naughty and nice for over a billion children. The North Pole collects structured and unstructured data ranging from customer profiles and gift-package radio frequency identification (RFID) tags to Offline Sibling Sentiment Data (OSSD) into a single Naughty/Nice Child Data (NNCD) database in real time that allows Santa align children’s behavior and requests with appropriate gifts.

Nope, nothing creepy about that!

The below infographic summarizes the data structure, and is the xAnalytica Christmas Graphic for 2015:
Runner up was Huskie’s analysis of the best days to go Christmas shopping in Dublin. xAnalytica’s Christmas Graphics from previous years:

Happy Holidays!


  • Comments Closed

World’s Most Accurate Pie Chart

Posted by – September 1, 2015

I am not a fan of pie charts, which to me add no understanding, but this is one that works:
Via I F**king Love Science, with thanks to Ed for the tip!

Unleashing Climate Data

Posted by – June 25, 2015

Freight tonnage by various modes of transportation, 2007

On June 17, the White House released over 90 data sets to help the transportation sector account for the impacts of climate change when examining and developing U.S. infrastructure. These include case studies, visualizations, tools and a compilation of key reports and websites from across the Federal government. One very well done example is above, a map of tonnage shipped by various modes of transportation, where the thickness of the lines varies by tonnage. Interesting to see how important the Mississippi and Ohio Rivers are for transport, and how much oil apparently is shipped via the Kinder Morgan Express from South Dakota to pipelines in Missouri and Illinois.

Thanks to AAAS for the tip!


Visualizing Hurricane Track Uncertainty

Posted by – April 19, 2015

The National Hurricane Center has steadily improved hurricane models over the past 25 years, as shown in this animation of uncertainties in the 48-hour predicted location of a hurricane’s center (mp4 video):
NHC 48 hr forecast uncertainty

NHC is also prototyping
storm surge mapping tools and conducting in-house (non-public) experiments with extending tropical cyclone track and intensity forecasts to seven days from the current five-day period; creating track and intensity forecasts for disturbances with a high chance of formation; and issuing tropical cyclone watches and warnings before cyclones form.


P.S. Although I like seeing how uncertainty declines with time, I don’t think the animation adds much to this over the simple contours labelled by year.

Virtual Trek

Posted by – March 14, 2015

This week Google announced it had collaborated with accomplished Nepalese mountaineer Apa Sherpa to create a virtual trek through Nepal using Google Map’s Streetview. Apa Sherpa, who has summited Mt. Everest 20 times, spent 10 days with Google last year to capture the images, which are now available on Google Maps. The effect is stunning, allowing you to virtually hike through valleys and towns near Mt. Everest, such as Thame, Khumjung and Phortse.

And for those of you hoping for a virtual tour of that other Trek, you might try here or here.


Bitter Christmas

Posted by – December 20, 2014

The HappyPlace blog at eCards provides the Christmas Graphic for 2014:

Just Who is This Guy?

And remember: always leave cookies out for the stoned burgler; the crumbs provide DNA evidence. Want more? Well, here are xAnalytica’s Christmas Graphics from previous years:

Happy Holidays!


Best Map App

Posted by – November 16, 2014

Google Maps is still the gold standard for smartphone mapping apps, according to NY Times columnist Molly Woods, who recently compared three smartphone mapping applications – Google Maps (iOS, Android), Apple Maps (iOS), and Here (WinPhone, Nokia). Apple Maps has made progress since its failed 2012 debut, but, according to Woods, Google Maps is still the winner. All of these provide some version of turn-by-turn directions, voice integration, flyover and vector-based maps; given that they are all free, I’d suggest you try out whichever one is available for your smartphone to see which is best for you. My choice is None of the Above: I’ve used Scout, (free, iOS) for the past two years because I like its interface and its ability to share my ETA along with a link to a map of my current location. And, if you subscribe, you can download the OpenStreetMaps region of your choice and use Scout off-line – a real plus if you are out of cell range or during disaster responses.



Posted by – June 17, 2014

One website I visit repeatedly is the New York Times’ Upshot blog, edited by David Leonhardt, because I find its graphical analysis of politics and economics help me think and understand. For example, they take a look at the monthly employment statistics that are often cited as economic indicators, and show that the typical level of randomness in monthly jobs numbers masks the underlying trend. They illustrate this using an animated simulation of random errors added to an average or a trend – i.e., a Monte Carlo simulation of jobs numbers – showing how easy it is to be fooled by randomness. The Upshot’s interactive forecasting model of the upcoming Senatorial elections, created by Amanda Cox and Josh Katz, is also worth a look.

Monte Carlo Simulation of Jobs data