Flourish Studios Sunburst Chart Tutorial & Pradeep Kumar G has our Viz of the Week
A Tableau-centric weekly blog about the viz-making process, #datafam member interviews, Viz of the Week & entertainment for introverts (consisting of a binge bite & a music morsel).
This week we discuss a data viz I (quickly) developed with a snazzy online visualization tool called Flourish Studios + Pradeep Kumar G (@pradeep_zen) shares powerful insights for our Viz of the Week (a viz made for the #TheSDGVizProject covered in last week’s blog).
Viz Process Loading…
Yes, this blog is predominately about Tableau. The tool and the community have done so much for me, I can never friend-zone it — yet, there are times my eye wanders a bit.
For example, the Week 7 2020 #MakeoverMonday dataset had three columns of data based on a country/region’s net worth. The fields are: ‘Country’, ‘Region’ and ‘Wealth’ (in billions, U.S. currency). There were 214 rows (mostly countries, but some inconsistencies and null values needed to be purged). Seeing this, the dataset spoke to me (well… actually whined and begged like data toddlers do).
I want to be sunburst[ed]. Please, pretty please!
After a few times of that, the (not so) dulcet tones of the data reverberated in my weathered skull.
Before I complied with his request, I wanted to see a couple things. With a sunburst chart, you need a hierarchical relationship (generally two dimensional fields for a hierarchy … otherwise it can be a bit cluttered — think category and sub-category). Check the graphic I created below to get a glimpse of the data structure.
Next you need a value to correspond with those fields. We had that… so far, so good. The next thing is to check the number of categories (i.e. the hierarchical queens) — this dataset had seven. It’s a bit much and kind of pushed it for the chart type, but we were still in the game. Another rub, we had 200+ countries (or sub-categories). This is trouble, Mr. Dataset. How can someone gain a bit of insight from so many slices and sliced slices? (1)
In the past, I used Tableau Magic (@Tableau_Magic)/Toan Hoang’s (@thoang1000) sunburst chart successfully as the ‘final’ exam of his wonderful Udemy course (Bespoke Data Visualisations in Tableau) and even cracked open Tableau to begin to re-form the dataset to comply with creating the sunburst chart in Tableau, but then I remembered Flourish Studio has Sunburst Charts out-of-the-box and probably animat-able. Groovy!
I popped in and sure enough it IS animatable! All I needed to was to define fields to sunburst, style and presto. Mr. Dataset was super giddy and I finished the viz in record time.
What are you saying? You don’t need to use trig, polygons and other complex stuff?
Nope, not at all.
So why did you bring up structuring the data, hierarchies and stuff?
It helps to provide context, so you can figure out the basics of how this is built — as a result, creating this chart in Flourish Studios is somewhat informed and doesn’t get made because of a happy accident.
Just because I’m in a good mood, I’ll provide a tutorial on the steps I made in Flourish Studios, so you can make like a tree and leave (after finishing this blog) to create this in five minutes (or better yet, your own version).
Creating a visualization is so simple, it’s definitely a cheat code! The cheat code is that their templates make the viz for you, but to help you understand the ‘point and click’ application, I covered many items. However, you can probably get a good start on a viz in the time it takes you to read the following…
Step 1: Find your chart from dozens of examples and click
For this example, I used ‘Sunburst’.
Step 2: Select ‘Sunburst’ and then change ‘Preview’ to ‘Data’ (as seen below)
It defaults to the ‘Preview’, but to insert your data and come to grips with the structure, you should always start with ‘Data’.
Step 3: Verify or restructure your data for the sunburst
The data structure is laid out for you using their example. If your data is not set up the same way, it’s an easy way to verify how it needs to be modified to meet the need of this chart. Basically, you need a hierarchical relationship with at least two dimensions and a corresponding value as explained above. Review & select columns to visualize (Note: they even color-code this for you!).
Step 4: Once you verify the structure of your data, upload your data file
Step 5: Define the ‘Select Columns to Visualize’ to visualize
- You will want the hierarchy queen on the inside and each lower level above that, so for mine… ‘Region’ is inside and ‘Country’ is outside (or ‘Region’, ‘Country’
- Size by is going to be your value field (‘Wealth’) because you want your values to determine each slice size (like a pie chart, donut chart or treemap to name a few)
- You can filter if you desire, but animating the sunburst effectively filters
- Finally, we have ‘info for pop-ups’, which is similar to tool tips for Tableau users and I used my ‘Wealth’ field also
→ Tip: Check the little visual on the bottom, it’s a keen little preview to show you how your viz will look without needing to access the ‘Preview’ tab
Step 6: Modifying the visualization via ‘Preview’
Preview acts sort of like Tableau Desktop, but the formatting toolbar for the viz is always available. From ‘Preview’ you can easily see how your viz will act once it’s saved.
Step 7: Understanding and working with the option descriptions (from the Preview Pane)
Layout — You can change the layout/chart type from here based on the series of templates selected. I would caution against this in many cases because the data often needs to be organized differently for other chart types to be effective. Other bits…
- Visible Levels — The levels will likely default to what was applied to the data tab, so you should not need to manipulate that figure
- Size By — You will almost always want to use sum in an instance with a value attached
- Auto Set Height — Defaults to selected and will probably choose to keep it that way in most situations
Labels — This section can be used to size, make ‘all caps’ or modify weight your labels.
Sunburst — With the ‘Sunburst’ section, the inner radius, depth fade and label rotation can be managed.
Colours — The Colours section is a lot of fun and really can offer a little character to your chart.
- Scheme — You can choose ‘Categorical’, ‘Sequential’ or ‘Diverging’. With a sunburst chart, it defaults to categorical and is generally the way to go with this chart type
- Palette — Flourish Studios comes with selectable palettes, which I used here (‘Antique’) , but a person can also override by color name, hex code or rgb to use their custom own palette
- Auto-extend — Is an option that defaults to prevent colors being used multiple times, but can be turned off
- Zoom out items and labels — These items are also colorable, but keep in mind the color of your dashboard and consider contrast
Pop-Ups — These work as your tooltips. It defaults to pointer/shadow to provide a 3-D effect, but you can customize to appear how you desire (similar to a rich text editor). You also have the ability to modify how numbers appear in the pop-up.
Controls — Controls mostly apply to work with animations are formattable with color and padding options.
Animations — Animations are customizable with this particular chart using duration (in seconds) of the animation.
Numbers — Include formatting options for decimals and thousands and a sample display
Font — Shares the limited number of font styles (currently eight) that could be applied to the visualization — you can also set a maximum width (but defaults with a blank to appropriately size with the viz)
Layout — Applies the color, several layout options and spacing applicable to the viz.
Header — Here you can add, format the visualization’s title, subtitle and additional text; another feature here is you can add a linkable image/logo for the visualization.
Footer — Used to add source(s) and their links + additional viz notes with formatting options
Step 8: Congratulations, now you can publish you viz!
You completed your visualization, but want to check it out one more time using various devises before publishing. Flourish Studios has you covered here as well.
Immediately below your name on the upper right hand corner is a preview bar. This allows you to preview full screen, desktop, tablet, phone or custom size. For me, allowing Flourish Studios auto-size works well on all devices, but would very extremely useful if you custom sized layouts, fonts and whatnot.
Once everything looks great, just save and publish… for free!
Bonus Tip: Adding your Flourish Studios creation to your Tableau visualization
You knew I would incorporate Tableau in here somewhere. If you really wanted to embed, it’s very simple; here’s a viz I applied it to Tableau.
- Create a web page part on your dashboard
- Paste your public Flourish Studios viz in the URL section. It should resemble: ‘https://public.flourish.studio/visualisation/9999999/’
- Size appropriately to this viz. Note: If you do not like how it looks on Tableau, you can always go back to Flourish Studios project and test there.
Thoughts & Considerations
- Personally, I would love for a complete revamp the entire Tableau ‘Show Me’ section to a big template focused area with various chart types and so forth similar to what Flourish Studios does so effectively. This approach would make it much easier for new users to learn/reverse engineer various types of charting on their own.
- Sunburst charts are almost never best practice. It’s basically a less insightful combo of a donut chart and treemap chart. There are so many ways to get more immediate insight. A couple that come to mind are bar charts, reducing the number of countries, comparing wealth and population to get a net worth per person and so many more. Make this chart ‘for the fun of it’. With that, I strongly believe that when animations are used more frequently in data visualizations, it will open up a larger world of what is an ‘acceptable’ viz — especially since the custom visualization can be used as an immediate attention grabber to begin the data quest.
- The designing of the actual dashboard in Flourish Studios has some limitations. There are so many cool options and customizations (as seen in the tutorial), but you can’t help but feel the loss of being able to manipulate objects on a dashboard to add more context to the data.
- If you really desire to understand how to make a custom chart, ‘creating’ one in Flourish Studios isn’t a substitute for building it by hand (2). When animations in Tableau are publicly available, I plan to build the vizzes to their animations, but may test-drive the charting and animations in Flourish 1st (as sort of wire frame or proof of ‘some’ concept) and hope to build and recreate in Tableau.
Viz of the Week
Adam Mico (AM): It’s great that you’re back to vizzing on Tableau Public. Your ‘Poverty’ viz is a powerful submission for the #TheSDGVizProject. How did you approach this design and were their attempts that didn’t go as planned before you found this beautiful iteration?
Pradeep Kumar G (PK): Yeah, I was out for a bit because I was relocating to Dubai, United Arab Emirates to join Beinex onshore team. After things settled down a bit, I was able to find some time to get back into the action.
I was very excited to learn about #TheSDGVizProject initiative [by Brian Moore (@BMooreWasTaken) Jacqui Moore (@jaxx084) & Vinodh Kumar V R (@VinodhDataArt)] as it deals with Global Goals and I love to create vizzes that spread awareness about global issues.
After the first dataset (Goal 1 — No Poverty) of #TheSDGVizProject was revealed, I spent some time exploring the data. The data contains information related to causation of poverty and decided to concentrate on specific reasons. When it came to the design, I opted to utilize the red color theme as it matched the logo. Instead of providing normal headers at the very top, I decided to show a map on the left with the header information on the right.
At first, I planned to create one view that showed the % of the population below the poverty line (only the map portion). Then, I decided to add some additional reasons for poverty into the viz to support the storyline. I segmented the viz into sections for three visuals and added BANs to highlight percentages and important figures. I added greyed-out icons (created in PowerPoint) (3) for each section to make it easy for the audience to identify and understand.
As I typically do, once I’ve created the general framework of the viz and the storyline, I start experimenting with different color themes, chart types, layouts, and graphic elements (such as icons). I also try to ensure that my viz is not cluttered. Instead, I try to implement a simple and elegant viz style, which are inspired by the amazing vizzes on Tableau Public. Thanks to #datafam. :)
AM: Even more than the design itself is the knowledge nuggets the data promotes. When you dug into the data, how did you know which data points to highlight and were there any more data bits you simply wanted to add, but could not?
PK: I tried to utilize the visual aspects in a way that they supported the storyline and made the insights to pop out of the viz. When I created a map showing the % of the population living below the national poverty line, I found out that in Africa, on average, nearly 42% of the population is below the poverty line; that was quite shocking! I was curious to know the reasons behind this sad truth. I then started looking at the dataset in-depth to find other reasons for Africa’s poverty. I learned that most people don’t even have access to freshwater.
I feel it would have been better if population and income data could be added to bring more insights. Though my intention was to represent a big picture of poverty statistics across the globe.
AM: It seems like the story elements of the visualization are tied together by items that could impact poverty including trends and locations impacted. Although there is some good news (e.g. worldwide trends), it appears many places, particularly in Africa, are having severe restrictions to basic drinking water services. Why did you feel this was an important point, besides income to highlight poverty?
PK: Water is essential for human life and access to safe water is a fundamental human need. Maslow’s Hierarchy of Needs also represents this well.
When looking into the dataset, particularly in the Democratic Republic of the Congo (DRC), nearly 45% of the population has access to basic drinking water services.
Below is one of the LinkedIn comments that I have received in feedback of my viz. It makes it clear that we can have an impact!
“That’s a great viz Pradeep! Despite their largest source of freshwater in Africa, The Democratic Republic of Congo also has the world’s largest deposit of cobalt and other significant mineral deposits like gold and diamond. This makes the country not only the richest in the African continent but potentially the world’s richest country. It’s quite unbelievable how this country is highly underdeveloped.”
I recently noticed ‘Purple Rain’ is on Netflix! For those scoring at home, the music is 11/10, nostalgia is 10/10, story is a 6/10 and acting maybe a 3/10… so overall a 7.5 flick and weighted to a 9.5 because the music and nostalgia is the only reason to revisit ‘Purple Rain’ (the movie). During this period, purple pop perfection oozed so much from Prince he was handing off music to many artists (including the ‘competition’ in the movie).
Some songs will always be bops and this one without exception. It may linger into guilty pleasure territory, but who cares… I’ll risk subconsciously singing ‘no money man will win my love’ out loud.
1) In layperson’s terms, a sunburst is a super-fancy hybrid of a treemap and donut charts)
2) Even when using a helping hand like a tutorial.
3) Here is Kevin Flerlalge’s excellent tutorial on applying PowerPoint to Tableau.
4) Personal Notes
* I had a fun time co-presenting a Viz for Good and meeting Tableau Ambassador Vince Baumel in Madison for the Madison TUG event on 2/21. Vince is such a natural and inspirational public speaker and his presentation is a must for finding inspiration in our community — and despite his talent and accolades, he is also an incredibly kind and humble person.
* Last week, my health metrics were: