Cincinnati

The official link for the event is located at this url:

http://www.tableausoftware.com/learn/usergroups/cincinnati-user-group/2015-01-20

Please make sure you sign up if you are attending! There is limited capacity set on the event posting, if the event fills up, I'll adjust the capacity to get as many people as we can fit into the space.  There is a paid parking garage below the building, street parking I believe goes till 9PM (so watch out for meters if you park on the street).

If you have any questions/comments/concerns, please reply to this posting or contact me directly at russellspangler <at> gmail.com.

Thanks.

Sure.

Prior to the competition multiple teams will need to form and register, those can be either individuals or groups.  I will release a random sample of the data about 2-4 weeks before March 19th. During that time the teams can practice on creating a viz within Tableau.

For the competition, each team will need to bring a laptop (only one laptop can be used for each team, however people can swap who is creating the viz if multiple people are on a team) and we also hope to have projectors for each of the teams.  At the start of the competition each team will only be allowed to have a blank workbook and the data set loaded to their laptops.  Each team will then have 30 minutes to build a visualization to present to the judges (they can also use other programs, such as Excel, R, Websites, or SQL, however all code must be written in the 30 minutes).  Visualizations can be single dashboards, multiple tabs, story books, etc. and use any kind of charts, along with focus on any part of the data (maybe someone does a sales viz while another team focuses on employee performance). It is up to each team to come up with an idea of the “best viz” for the data. 

We’ll try and provide commentary as the contest progresses during the 30 minutes.

After the 30 minutes have ended, each team will have a few minutes to present/talk about what they built and why.  Then the judges will ask a few questions or make a few suggestions and write down their scores based on a few categories (example: Visualization Design, Creativeness, Overall (TBD, I’ll provide more details about points in each category and what judges will be looking for)).  We’ll total up the points given by the judges and announce the winners, and we might even do something to determine a crowd favorite.

For those not in the competition, you can watch multiple people work on building a viz and hopefully gain some tips and tricks in the process.  As always after the event is over I encourage people to socialize and hopefully meet others in the area (especially if you enjoyed the viz that one team built).

We’ll also try and post all the vizzes online to the user group page/Tableau public profiles and will help support the teams/individuals via social media.

That is the general plan and I'm open to making some changes if needed.

Kevin, I'm not sure if anyone answered your questions, but I'm happy to add my 2 cents.

I normally would suggest having your single point of truth within your database, that way you can connect multiple tools to one point within your organization.  Even though you might build something in Excel and something in Tableau, having data staged in a database will allow you a place to pull the exact same raw data for either of the tools.  If you do that you can avoid having published data sources and have a data connection directly to your backend.  You can still do published sources and have them refresh on an automatic schedule.  Granted if you use multiple tools you still have the issue of building things incorrectly and having mismatching numbers.

If you refresh a published data source it will refresh the data source on the server, if multiple workbooks connect to that published data source, you are correct that multiple workbooks will be updated.

You can do one production server with a folder for development.  Most development should happen within the desktop product, once a workbook has been put together it can be pushed to the server for final testing. If you don't want to manage multiple environments, having a folder for development might provide the least amount of maintenance.  If you have a large team and critical reports, then having two separate server environments and only a select few members to do the publishing to the production server might provide you with the most control (you can also do this separation using sites).

The hierarchy of Tableau is: Server -> Sites -> Projects -> Workbooks -> Worksheets/dashboards.  Sites are separated on the server (however users can access different sites on the same server), projects are like folders (only 1 level deep), and workbooks and worksheets/dashboards are what you see when using the developer product.

I hope that helps!

Great job last night, guys! Can't wait for the next one.

The Iron Viz against the other Local Groups seems like it will be lots of fun. And I am sure we will win, with all the talent we have around here.

Also, from the discussion, here is some info about hiding underlying data:

http://kb.tableau.com/articles/howto/hiding-underlying-data

Question

How to hide the underlying data when sharing a workbook.

Tableau Desktop

Use one of the following workarounds:

• Publish the workbook to Tableau Server, and deny permission to view underlying data to relevant users. For more information, see Publishing to Tableau Server in Tableau Help.
• Hide unused fields, and then create an extract. Hidden fields are automatically excluded from the extract. For more information, see Hide or Unhide Fields and Extracting Data in Tableau Help.
• Use an aggregated extract. Aggregating data prevents users from seeing row-level data for aggregated fields. For more information, see Aggregated Extracts in the Tableau knowledge base.

• When using Tableau Reader, you cannot access the underlying data.
• To voice your support for the inclusion of this feature in a future product release, add your vote to the following Community Idea: Allow download of summary but not underlying data

Click on the option on the right side of any pill and click format. That should allow you to format the number to a %. If you do it on the canvas (rows or columns or marks) then it only holds for that use. If you do it on the data panel then it formats universal for all uses.

Or in the data pane.