For the 2016 Tableau Conference in Austin, and I have unified our previously separate work on building Twitter network graphs in Tableau.

Incorporating text analytics, our aim was to update the view at steady increments throughout the conference.

You can find our earlier pieces on Tableau Public at these links:

And here is the :

Project Wrap-Up

Chris has published his write-up about the project . For my retrospective, I will highlight aspects of the data pipeline, the tool sets, and the collaboration.

Vectorization

Various pre-compute steps were executed independently within the overall workflow for each topic:

• keyword parsing (Python)
• keyword scoring (Python)
• network coordinates generation (R)
• network centrality measurements (R)
• orchestration & data reshaping (Alteryx)

So, with 28 topics, you can imagine that I didn't want to run these five steps manually, for each topic on every data refresh! So vectorizing these individual components inside of the overarching workflow was important for automation.

Multi Disciplinary

Making use of four tools, Python > Alteryx > R > Tableau, our pipeline was rather sophisticated.

Each tool has an inherent strength, and it follows naturally that all four analytics environments had

Detailing Twitter mentions from across four years of the annual Tableau Conference, in a collection of 45 interactive network graphs, this project is published in close collaboration with . He is also presenting a curated collection of his beautiful hive plots from the same data.

You can find our two pieces on Tableau Public at these links:

Bringing It Together

My interest in the analysis of network graphs first piqued while studying in Stanford Online MOOC, . A graduate level course intensive in math and theory, it was challenging; and also left me wanting for real world application of the concepts I had learned.

Bringing together my recent studies in R, Alteryx and Tableau, this project is that application.

If public data from Twitter is perhaps relatively benign? Then consider the power of enabling visual exploration of other more highly valued network

This post imbues the importance of innovation with color in data visualization, offers a variety of resources and reference materials, and encourages personal innovation with color as absolutely vital to moving your visual communication of data forward in Tableau.

Emotion and Behavior

The effective use of color is fundamental to the visual communication of data.

As our eyes take in color, they communicate with the hypothalamus, which in turn signals the pituitary gland. Then, on to the endocrine system, the thyroid gland signals the release of hormones. Those hormones influence BEHAVIORS and EMOTIONS. Color is so powerful, in fact, that the effective use of color can improve learning by 75% and increase comprehension by up to 73%.

Yet, in today's conversation about color, much ado is still invested in the basics: to , for example.

Important as these basics are, now is the time to move our conversation beyond the entry level. Now is the time to dramatically expand our thinking around color.

With behavior change, comprehension, and augmented decisioning as the purpose of data visualization, and Tableau as our tool of choice in the field, then we as visualization authors must become more sophisticated

This post builds upon the theme of designing a performant data architecture for your high volume solutions in Tableau.

One core performance concept is that good design considers the entire solution stack. If you fail to design for performance at all vertical levels, then the worst performing layer will make the solution slow. A train is only as fast as the slowest car. And worse, if various layers have design problems, then your train likely isn’t moving at all.

We must consider the entire vertical solution, together as a holistic system, from the top to the bottom. And this design investment is best made at the outset. To focus performance efforts at only a single layer or to return to a poor performing system in hindsight in search of "one thing" to fix is insufficient.

Of the various layers in the typical solution stack, this post is focused on two: User Interface Design and Semantic Data Architecture.

Yes, as UI designer in Tableau you are also a Data Architect!

Guided Analysis

As with any “big problem", the solution to good performance on large data volumes is to break that big problem up

This post describes how an "advanced menu" can be used to work around the need for dynamic parameters when filtering across multiple data sources in Tableau.

Concept

In his post "Creating a Collapsing Menu Container in Tableau", Robert Rouse does a great job of walking through the mechanics of how to build a "dynamic and collapsable menu" in Tableau.

Some of my favorite mobile apps like Slack, Feedly and Google Maps have a slide-out menu that appears when I tap a small icon. That common design element makes plenty of room for user inputs and gets them out of the way when you're done - perfect for small screens.

To elaborate further on that concept, in this note today I explain how we can leverage the idea of a "dynamic and collapsable menu” to tackle some additional, rather complex data design challenges.

Why Multiple Data Sources?

First off, why would we deliberately use multiple data sources in a single dashboard?

Well, on large data volumes, for performance! In fact as your data volume grows large, Data Architecture decisions like this one quickly become imperative.

For two years running at the annual Tableau Conference,

Here you'll find an example to build upon when writing to those who insist on using Tableau to build "giant walls of tabular data".

...

Dear person who insists on giant cross-tabs of data in Tableau:

It was a pleasure to speak this morning! As we’ve discussed last week, the difficulties you face stem from the fact that you are attempting to do with Tableau what is specifically not recommended.

Tableau is a data visualization tool. It is not a spreadsheet, not a “tabular report builder”.

After looking at your challenges in more detail, it would seem that you must speak with your stakeholders and soon decide between one of two broad categories of alternatives:

Choice #1: continue to use spreadsheets and "giant walls of raw numbers with conditional formatting" to make business decisions

• Here, your best decision may likely be to avoid Tableau

Choice #2: leverage the visual display of quantitative information to enhance cognition and reach better business conclusions faster!

• Here, continue with Tableau and render your data visually

In support of the above conclusion, please find below a collection of reading materials.

What is Tableau not really good for?

"We

Bolstered by the brain trust at , this post considers various uses for in Tableau, and argues for more formal data preparation as the best alternative when blending breaks down.

For Starters

If you're just getting started, first some useful resources:

• &

All of the 2014 conference materials are an excellent resource. There are ten different talks with the keyword “blending", and my makes it easy to find what you’re looking for.

So now, on with the show!

Slide Projector

As an analogy, think of Tableau as a slide projector for your data where each Tableau Data Source is a slide.

Born from a hackathon among Tableau’s engineers, Data Blending is indeed a clever hack! It allows us to place more than one slide into the projector at once :)

Starting in version 8, "Data Blending 2” also allows us to manually turn off & on the linking fields, regardless of whether those fields are utilized in the view. The difference

This post helps you to understand how the granularity & the shape of your underlying data will affect the visualization work you do in Tableau.

I was recently stumped by a common problem, one for which Jonathan Drummey back in 2012 has started a page to document the many scenarios in which this type of complication can occur.

My particular scenario was . And Jonathan's collection of similar scenarios is

The most interesting thing about each of the scenarios in Jonathan's collection is not their individual solution in isolation. But rather, the underlying pattern behind those solutions: the what they share in common.

And when Joe Mako helps me get through something on a Sunday afternoon, you know the answer is worth sharing!

The number one, most important facet of learning Tableau, and learning from Joe, is to recognize the patterns that recur. By recognizing common patterns when working with data, and by learning the behaviors of Tableau, one learns to reach a flow state with similar encounters in the future, even while the details may vary.