Another Round of 6 Techniques

By Ross Perez 2010/10/10

The questions keep flooding in, with people clamoring for more great 6 techniques. No problem though; we’re here to ensure a great 6 experience.

Michelle writes: Is there any way to do a percent of total calculation per row and then utilize that calculation on the page (i.e., color coding or sorting by the value)?

Michelle, we know that attention to detail is crucial for some people when it comes to 6, so thanks for such a good question! Here’s the short answer: Yes! And the longer version: Click on the drop down on the measure you are looking to calculate, and choose Quick Table calculation-> Percent of Total. Then you can edit it (again through that same menu, but choose edit table calculation) to calculate at whatever level you please. You can then use CTRL to copy that field, or drag and drop it to the color shelf to color code.

How can I get the map graph to show by city and state versus longitude/latitude - or is that possible?

This is definitely possible. Right click on the field you are looking to Geocode, go to the menu item Geographic Role and specify how you would like it to geocode. Then, double click on it (on a new worksheet) and it will display it for you over a map. Need more details? See our KnowledgeBase; this article in particular shares map basics.

How does the "optimize extract" feature work?

This is a question that could come only from the truly data-curious! The feature works by compressing your data and placing it into a format that can be queried quickly by Tableau. Perhaps most importantly, the queries that we do send to the extract automatically configure themselves to your hard disk, RAM and cache memory to take advantage of all of their strengths and avoid all of their weaknesses. That’s the short version—please contact Customer Support if you need more details.

What is on the list for 7.0?

Right now it’s hard to believe anything could be better than 6, but when the time comes for a new upgrade, it will assuredly have even more of the amazing and useful features you’ve come to expect from Tableau. We love your feedback and ideas so send them along.

Christopher wonders: One of the struggles/frustrations I have had with Tableau up until now was its inability to count unique records. … I understand that Tableau 6 enables me to accomplish this. What do I need to do with my data to enable this?

As long as you are using the 6y new Data Engine, you don’t need to do anything special to your data to accomplish this. Just right-click drag a dimension to the view, or use the drop down in the pill to change the aggregation to Count Distinct.

Next Question – I have multiple workbooks that I have connected to a single Access database table. If I change all of my connections to map to the new Data Engine generated source; how do I update my data?

Sometimes the tried-and-true techniques are the best ones. You can update your data the same way you used to. Here’s the process:

  • You have multiple files connected to the same extract, which is based on an Access database.
  • Open any workbook that accesses that extract (a .tde file). It will ask you to upgrade your extract. You should choose yes. This will change your extract to a Tableau 6.0 Data Engine extract but will not update the data from your Access database.
  • If you wish to refresh the data in the extract, go to Data menu -> Extract -> Refresh. This will cause your extract to update with the latest from your Access database.
  • Open any other files that access the same extract (same .tde file) and make sure they are working. You can check that you are accessing the right extract file by going to Data menu -> Extract -> Properties. The Database field tells you which extract you are using. You should not need to switch to live connection and back.

Ijaz asks: With all the speed that has been touted using 6, is that utilizing the Tableau extract?

You will experience speed improvements in every aspect of Tableau 6. The Data Engine (extracts) are the area with the most massive improvement, on the order of 100x. However, every part of your Tableau process from connecting to SQL Server to publishing to Server and serving a live view should be improved.

Phew! That was quite a 6 marathon with Michelle, Christopher and Ijaz, but once we rehydrate we’ll be game for more. Keep those questions coming!