Google Online Takedown Requests: a Lesson in Extract Aggregations

Tis the season of visualization contests.

I've entered this viz in the Visualizing Online Takedown Contest at Visualizing.org.  You should check out the submissions, there's some truly inspiring work there.

As you can imagine, this was a large chunk of data (600k+ rows) for Tableau Public, which restricts rows to 100k.  Being used to using massive amounts of data in Tableau Professional, I of course forgot about this restriction until I went to publish it to the web on the morning of the contest submission.

In order to publish it I had to aggregate the data to month from day/minute.  Unfortunately, I had presented the data at the week level, which is not an option for extract aggregation, so I also had to change some of -the views.  Note to self:  add this to the Ideas on Tableau Community.

To do this:  

Look at your data first.  The aggregation will roll-up your measures based on unique dimensions, so if you have a unique id field for every row, that won't be rolling up.  With this data, I replaced the Request ID field with the value of 1, that way it summed up to the number of  requests for a Copyright owner for the month.  You should also note if there are pre-calculated metric values (rates, percentages), as you don't want those to aggregate.

Right click on the data source and select Extract Data

Select the appropriate aggregation level and Extract

Voila!  Tableau brilliantly reduced the records from 600k+ to 40k.  Sweet feature!  Have I mentioned lately how much I love this program?

I should also mention that the bubbles view is built using the index() function.  For more info on how to do this, check out Steve Wexler's blog post:

My Four Favorite (and Underused) Visualization Types

No comments:

Post a Comment