A reasonable approximation would be to sort states according to the percent of time they voted for a given party. To break ties we can sort by the latest year that each state voted for that party. Here's the resulting image:

The sort order is based on party preference aggregated over time, which is difficult to perform in the view above since time itself is a dimension; we must first sort in the absence of time, and allow that sort order to persist as we evolve the visualization. To create the view above let's start with a simple view of party vs. state.

Next we define a calculated field 'CntPartyLastYear' that will aggregate the count of a state's party preference over time, along with the tie-breaking value of the most recent year a state voted for each party (converted to a fractional value).

• CntPartyLastYear:  COUNT([Party]) + MAX([Year]) / 10000

Place this field on the "Text" shelf and create a table calculation for the percent of a state's total vote to see the split in party preference aggregated over time:

Finally select a column (I arbitrarily chose the Republican party) and sort.  This sort order will persist as we modify the visualization to create the final image.  Remove 'CntPartyLastYear' from the Text shelf.  Move 'Party' from the Columns shelf to the Color shelf, and place 'Year' (discrete, all values) on the Columns shelf.  Note that this final step now incorporates time without disrupting our sort based on aggregation over time.

If you'd like to experiment, here's the packaged workbook (TWBX). We offer a free trial that you can quickly download and install to explore this data and much more.

The attached packaged workbook offers quick filters that let you interactively explore the leading/lagging correlation in more detail. You can view this packaged workbook using the free Tableau Reader, or for a complete experience download the free trial of Tableau Desktop.

This workbook helped us explore how to generate a smoothed version of the raw polling and Intrade data, as visualized in the middle pane of our guest post chart. This was a key step in our analysis, because the final pane uses a rate-of-change table calculation in Tableau that is very sensitive to the rapid day-to-day changes in both the polling data and the Intrade data. Like focusing a camera lens, we needed to blur the short-duration changes in order to focus our attention on the longer-term trends.

Using Tableau's quick filters in the attached workbook, you can experiment with the parameters for our approximate-Gaussian smoothing filter. You can see how important it is to use a good low-pass filter; for example, a Gaussian must be “clipped” in practice (to yield a finite support), but too much clipping erodes its low-pass characteristics and produces graphs which are significantly obscured by high-frequency noise.

Tableau has also been demonstrated in the Teradata kiosk that is focused on v13 features and functions. The demo uses 20 years of FAA data that is stored in a Teradata EDW. Tableau is able to employ the Teradata geo-spatial UDFs so that "where" questions can be easily asked. The results are placed on the maps that are automatically generated in the application. There have been many positive comments about the utility of Tableau's mapping capabilities -- especially around the ease of use.

In addition to all the terrific technology partnering at the conference, Lance Armstrong gave a compelling keynote yesterday morning. He shared his ordeal with cancer and described how he used it to re-focus his life and his career. The message was a real inspiration. There was also a little "French ribbing" which got some laughs...

Utah loves their data rockstars!
Utah delivered a lot of players and some big scores.

The most popular songs were the least enjoyable to listen to...
If I hear Roxanne or Should I Stay or Should I Go one more time, I'll need a lot more beer than we had in the booth.

Explore the rockstar data.

+ Download a 14-day Tableau Desktop trial -or- download free Tableau Reader.

+ Download the Rockstar packaged workbook. The file includes the views and data shown above. Open it up in either Tableau Desktop or Tableau Reader and have a jam session with Rock Band data.

You can see that despite the attention that the 2000 election brought to the electoral college, George W. Bush’s win was the 14th time that the ultimate winner earned less than 50% of the popular vote but more than 50% of the electoral college. Among the other thirteen times, 3rd-party candidates may have been the "spoilers" (as in Bill Clinton’s 1992 win over George H.W. Bush and Ross Perot). Nonetheless, it’s an interesting observation to realize how often this phenomenon has occurred.

Three interesting notes: John Adams (1796), Thomas Jefferson (1800) and John Q. Adams (1824) all earned a greater percentage of the popular vote than they did of the electoral college vote, with John Q. Adams not even hitting 50% of either. That was quite an election since it was eventually decided by the House of Representatives; for more details, see http://en.wikipedia.org/wiki/United_States_presidential_election,_1824.

Here is how you build this overlapping bar view of two measures in Tableau:

1. Build a bar view
   
2. Drag the second measure to the axis to combine it with the first measure
   
3. Move the Measure Names field to the color shelf
   
4. Turn off stacking
   
5. Adjust the overlap order by dragging items on the color shelf
   
6. Finally, write a calculated field to create an extra column to show when the electoral % was less than the popular %. Add it to the Measure Names/Values card.
   

Teradata's implementation of their Geospatial Extensions largely follows the SQL/MM standard. The ST_POINT datatype allows for two floating point values representing coordinates. To Tableau this appears as a binary data string, but Tableau's calculated fields can extract the longitude and latitude components using Teradata's user-defined methods for geospatial types. For example, given an ST_POINT column named AIRPORT_LOC we would create the following calculated fields named "lon" and "lat":

• lon: RAWSQL_REAL("%1.ST_X()", [AIRPORT_LOC])
• lat: RAWSQL_REAL("%1.ST_Y()", [AIRPORT_LOC])

Note that these are automatically recognized by Tableau as their appropriate geographic roles. To demonstrate Teradata’s Geospatial Extensions, we will explore a sample data set taken from twenty years of flight records. Teradata's user-defined method called 'sphericaldistance' performs great-circle calculations between two ST_POINT types. To determine the distance between an origin and destination (assuming a spherical path at sea level), we can create the following calculated field which includes a meters-to-miles conversion:

• distance: RAWSQL_REAL("%1.sphericaldistance(%2)", [ORIGIN_LOC], [DEST_LOC]) * 0.000621371192

With the distance computations we can explore a new aspect of the sample airline data. The view below examines delays by distance, in bins of 500 miles, and reveals an interesting spike in delays for flights between 3000 and 3500 miles. Interestingly, there are rarely weather delays at that distance.

Another quick way to view the flights in this category is a map of origin and destination points, filtered on distance.  The dashboard below helps isolate the delays to a handful of popular destinations, most notably Hawaii.

Tableau's calculated fields also allow us to construct UDTs that we can use in queries to the database.  For example, the plot below shows all airports within 100 miles of New York City.  The marks are size-encoded by the number of unique destinations served by each airport, and color-encoded by the average severity of delays. Of course, it’s no surprise that the three airports typically associated with NYC have high numbers of unique destinations; note that of the three, JFK has the highest severity of delays.

In order to create this view, we first need a calculated field representing New York City:

• New York City: RAWSQL_STR("NEW ST_POINT(-73.986951,40.756054)")

We can create a distance calculation and place it on the Filters shelf to limit our view to distances of less than 100 miles:

• Distance from NYC: RAWSQL_REAL("%1.sphericaldistance(%2)",[New York City],[ORIGIN_LOC])*0.000621371192

With fairly simple SQL expressions we can extend Tableau's support for new types of data and analysis on any database system that supports user-defined types, methods and functions.

Tableau has always believed that a fast data engine enhances the analytic experience. We have designed the Tableau products to leverage the power of the database engine to provide a live, always-current data analytic solution. There are a number of ways to get fast analytic queries. Dataupia, Netezza, and Teradata are all recent partners Tableau has extended our products to work with. These are databases specifically designed and optimized for high performance for analytic queries with large volumes of data. Microsoft has always been a part of our network with SQL Server and the high performance Microsoft Analysis Services (MSAS). Now Project Gemini promises to provide even greater performance to Analysis Services deployments, with even larger data volumes. Microsoft pre-briefed us last week and we applaud this project heartily.

With Project Gemini, Microsoft is building an in-memory data store as an option for Analysis Services. Once the data is loaded into the Microsoft in-memory data engine, a key aspect of the Project Gemini announcement is the publishing of this data to a SharePoint Server. By publishing the data to a server environment which includes SharePoint, Excel Server, and Analysis Services, the data becomes instantly available to all tools that can access the Analysis Services engine.

The new in-memory engine is in essence becoming an alternate storage mechanism for Analysis Services. Today, Analysis Services supports a MOLAP data engine (pre-built cubes of data) and a ROLAP data engine (data stored in relational format that is modeled to look like a cube). With Project Gemini, Analysis Services now gets a lightning fast, highly scalable in-memory data engine. And this will still look just like regular Analysis Services to any tool that uses MDX to connect to Analysis Services.

We believe there is no better tool to connect to Analysis Services than Tableau. We will simply inherit all the performance gains of the in-memory engine automatically with our support for MSAS today.

As a certified Microsoft Gold Partner, Tableau is excited about the new technology Microsoft is building. We’re looking forward to getting our hands on alpha & beta copies of it so we can start validating and further optimizing our support for the Microsoft data engine.

You can read more about Project Gemini and the Microsoft event.