June 29, 2016
This morning I was going through the backlog of Makeover Monday candidates that I have saved in Pocket and it’s quite an extensive list. So I thought I’d knock one off the list and start up #WTFWednesday for those of us that just can’t have enough Tableau in our lives.
What works well?
- It’s neatly organised by the amount of misconception
- Colors follow Guardian standards
- Nice title that sets the story
What doesn’t work well?
- When I first read it, I assumed the darker blue was the average guess, but that’s not the case. This should be made clearer.
- Comparing countries could be made easier
- Stacked bars with overlapping labels look very cluttered
I didn’t want to spend a lot of time on this so I decided to make a simple barbell chart sorted by the largest misconception. This was also the first time that I’ve put bar charts in tooltips, which works well for allowing the user to see the precise values.
June 28, 2016
This week’s tip shows you how to perfectly positive KPIs next to bars irrespective of the dimensions in the view. This started as a Data School Gym Challenge if you’d like to have a go at it before you watch the video.
June 27, 2016
Here’s another Data School Gym challenge for you. I’ll create a video tomorrow for how to solve it. First, let me explain the problem.
I’ve written before about how to use ASCII indicators to compare to a goal.
But there are times when we might want the indicator as part of the chart next to the bars. Start by creating a year over year comparison like this:
Then create a calculated field as a placeholder:
Add the Placeholder as the secondary axis and synchronize. Then create a calculation that compares the two years:
Change the mark type to shape for the Placeholder axis and add the comparison calcuation to the Shapes shelf and the Color shelf. Change the shape and color to something more meaningful.
Then right click on “True” on either the Shape or Color shelf and choose “Hide”. This cleans up the viz nicely. Go ahead and hide the secondary axis while you’re at it.
Ok, looks good. However, what happens when we add another dimension?
The indicator is fixed to -50,000 so now there’s a wider gap between the indicator and the bar (because the value of the primary axis is smaller).
Here’s the challenge:
Notice how in each of the screenshots below, the arrow is always precisely the same distance from the left side of the bar. That’s the tricky bit.
You can download the same version of Superstore Sales that I used for this example here.
What works well?
- It’s easy to see how many contributions have been made and how many people have contributed.
- The colors are good, but not great.
- The word cloud helps highlight the volume of people that have contributed.
- You can easily see which weeks had the best participation and which weeks drew the most new people.
- The heatmap on the 2nd tab does a good job of letting people see which weeks they have yet to complete.
What doesn’t work?
- It’s too hard to see the overall figures.
- The 2nd tab should be incorporated into the first.
- The word cloud is too busy.
- The cumulative graphs don’t provide much insight.
- All that’s important is the final number.
- I don’t like the combination chart on the upper left.
- There’s no interactivity.
I built my makeover in Tableau 10 so all I can post is an image. If you’d like to have a look at it, you can download it here. I’m really excited to see what the Community builds. I’m 99.9% sure someone will come up with something that blows us away.
June 26, 2016
So for week 26, we thought we’d have a little contest. I created a rather rudimentary dashboard to track the who, when and what of Makeover Monday, but I know you can come up with something better. For week 26, we’d like you to makeover the Makeover Monday dashboard.
Andy and I will review all of the and pick our favorite sometime this week. We’ll send that person some Information Lab and Tableau swag. I don’t know what we’ll send you yet, but that’s besides the point.
You can download the TDE here or XLS here. The data is as of 12:45pm Sunday 26 March. If I’m missing anything, tweet me and let me know. Good luck!
June 24, 2016
Today at the Data School the team was supposed to present Tableau Server adminstrator views, but given yesterday’s historic decision by the citizens of the U.K. to leave the EU, we pivoted and focused our day on visualisation data about the referendum.
I wanted to participate as well, and quickly found this survey on YouGov.UK. This poll was taken just before the vote, seemingly a good proxy for the next day’s vote.
June 22, 2016
I absolutely love this visualisation by FiveThirtyEight. It’s clean, uses minimal color, and is overall nice to look at.
So naturally I wanted to see if I could recreate this in Tableau, but more specifically Tableau 10. Here’s your challenge, reproduce my viz below. Post a link to your image in the comments or Tweet it to me. Download the TDE here or the Excel file here.
Here are some tips:
- Use Tableau 10 so that you can use the new Tableau fonts. Or just use the font of your choice if you don’t have Tableau 10.
- The fonts for the countries in black is 11pt while the font for the grey countries is 8pt.
- If you’re using Tableau 10, using groups in calculation will come in mighty handy!
And if you’re looking to do a bit more, here’s another version that only highlights the countries that would have an increased score. This one doesn’t require groups, but could certainly be done that way if you choose.
Good luck! Go smash it!
June 21, 2016
For this week’s Tableau Tip Tuesday I show you how to use Level of Detail expressions to get the latest year and prior year and then compare then to create visual alerts.
June 19, 2016
This past week felt like such a great week for learning and I feel fortunate that I have Makeover Monday to apply those learnings. There are two influences in particular I’d like to call out:
- Rob Radburn gave an amazing presentation about Makeover Monday at TC London. In his presentation he referenced the book “Steal Like an Artist” which I’ve read before. But then Rob went through several examples where he pointed out where he stole like an artist.
- The Data School gave presentations this week that focused on infographics. You can watch their presentations on YouTube here. I was totally blown away by their work and wanted to create one of my own this week.
Before we look at my viz, let’s have a quick look back at the original from Nippon.com.
What works well?
- The data is organized by the theft type, using color to distinguish the groups.
- The total number of thefts is included for context.
- Each outer ring is sorted properly, with “Other” being ranked last.
What doesn’t work well?
- A donut inside a donut is NEVER a good idea.
- Labeling every single slice makes the chart overwhelming and too busy.
- The theft types (inner donut) are in reverse order.
- The colors are ok, but not particularly outstanding.
- Making comparisons is very difficult.
As I mentioned, this week I wanted to create an infographic. Like Rob suggested, I did a quick Google search for inspiration and this identity theft chart from PC Magazine struck me for its colors and organization. Without further ado, here’s my Makeover Monday infographic.
I created this in Tableau 10; if you’d like to download it, you can do so here.
June 15, 2016
Today at Tableau on Tour London, I had the opportunity to share as many tips and tricks as I could in 30 minutes over lunch. For me, it’s super fun to share what I’ve learned from others, so thank you for the opportunity.
You can download the workbook here (requires Tableau 10). Below is the session recording. Enjoy!
June 14, 2016
This week’s Tableau Tip Tuesday is a little trick I figure out as a response to a question after Jesse Gebhardt’s session on dashboarding at TC London.
Can we allow a user to zoom in or out on an axis?
The answer is an emphatic YES INDEED! Watch the video to see how it’s done.
June 13, 2016
The chart for this week’s Makeover Monday is an excerpt from Women in the Workplace, a study undertaken by LeanIn.Org and McKinsey.
- The chart tells a powerful story of the decreasing representation of women as leadership levels increase.
- Using red for women draws the eye to those portions.
- The table allows you to compare to 2012.
- The pipes are ordered from lowest-level role to the highest-level role.
- Good reference to the source
What could be improved?
- They took the “pipeline” analogy too far by creating a "pipe chart".
- The pipes get smaller as you go left-to-right. I can only guess they are sized by the % of women in each role.
- The pipes are 3D.
- The pipes are filled up on an angle, making it way to difficult to gauge what percentage it represents.
- It’s nearly impossible to compare one role to another.
- The title could be more impactful.
- Do the pipes represent 2012 or 2015? There’s no way to tell.
- The chart doesn’t allow for comparisons between 2012 and 2015.
Basically, I think this chart is horrible. It’s quite possibly one of the worst I’ve seen in a long time. This isn’t about the author the person. Clearly they aren’t educated in data visualisation. I’d bet they working in marketing.
Again this week, I used Tableau’s story points feature to show my step-by-step makeover. This week was pretty quick. I was able to iterate through all of this in under an hour.
As a stand-alone version, I would make the chart slightly taller, like this one below. Click on it for an interactive version.
June 10, 2016
Alteryx Inspire was an amazing conference! And the name is absolutely spot on…you will be so inspired to do amazing things with Alteryx. While there, I decided to challenge myself and learn how to build an Alteryx app. My app, named Pinterest Board Downloader, allows you to download all of the information about any public Pinterest board. In this post, I’m going to detail how I built it, what I learned, and who helped me along the way.
A bit of background on where the app came from. I had been using IFTTT to log all of the pins I make to the Makeover Monday Pinterest board, but the main problem was that I didn’t have data for the first few weeks. Pinterest has an API that you can call to download all of the pins and Paul Houghton was teaching data parsing, data prep, JSON parsing, and macros in the first week of Alteryx training for cohort 3 of The Data School. I saw this as a great way for me to learn these skills too and also fix my data issues with Makeover Monday.
However, for the app, I wanted to make it more generic so that it would work for ANY Pinterest board. Let’s start with the guts of the workflow.
Essentially this is what the macro does:
- Take the input from the app (the URL of the Pinterest board) and call the Pinterest API
- Parse the JSON
- Clean up the JSON to make it wide rather than tall
- Interate through all pages of the Inboard (the API only returns a subset of the records with each call)
The parsing and cleaning bits were pretty straight forward and so was calling the API. What tripped me up was how to cycle through the pages in the Pinboard to get all records. That’s where Robin Kennedy showed me how to make it an interative macro. Make sure you choose the Iterative Macro option on the Workflow tab of the Configuration window.
I couldn’t quite get it to work though and the Data School presentations were looming. So I called Chris Love and he showed me how to make the macro page through the iterations. It’s like magic watching Chris use Alteryx.
Conceptually I understand what an interative macro is and what it does, but I still don’t have my head wrapped around WHY it works the way it does. I admit that I don’t quite “flow” with Alteryx yet like I do with Tableau.
Ok, so now I’m getting all of the records. The next step is to turn it into an app. I had never done this before, but it was fairly straight forward. The trick is to go to the Workflow tab on the Configuration window and choose the Analytic App option.
The app workflow is also pretty straight forward.
Andy Pick helped me get started with the interface design part. All it does is provide a text box for the user to enter the Pinterest board URL. From there, I update the API string to use my API key, but to get the data for the Pinboard entered. The app then calls the interative macro above and the Action tool asks the user where to save it and in what format. I didn’t want to only allow a CSV or TDE, so the user can save it in any Alteryx supported format.
After I got this bit working, the macro broke. This led me to the Community room that was set up at Inspire and there waiting for me was the great Chris Love. It turned out I wasn’t passing the string properly from my app into the macro. Once that was fixed, it worked perfectly.
And that’s it! I feel like I learn a massive amount at Inspire this week. The Community is amazing, enthusiastic and absolutely brilliant. The most important lesson I’d like to pass on is that in order to learn, you need to get stuck in. Find a real problem you’re trying to solve and see it through. Ask for help when you get stuck.
Give my Pinterest Board Downloader app a whirl! Send me any feedback if it isn’t working quite right.
June 7, 2016
This week’s Tableau Tip is about conditionally formatting discrete rows. Many of ushave used conditional formatting in Excel and we like how we can color-code our discrete items (e.g., customer name) based on a value that’s in the view. In this video, I show you how to do that in Tableau.
June 5, 2016
My process works like this:
- Evaluate the existing visualisation - What works well? What needs improvement?
- Rebuild the original chart in Tableau
- Capture each version of the chart along the way and indicate what I changed
- Finish with the final makeover
This process also helps me iterate quickly and keep as close to the one hour “recommended” time as possible. Without further ado, here’s my makeover for this week.
As I was about to land in San Diego for Alteryx Inspire I happened to look through my backlog of makeovers and saw this beauty.
The whole purpose of these charts is to show the difference between the tasks a data scientist spends time on and the least enjoyable part of their job. These two charts completely fail in telling getting that message across.
So with 15 minutes remaining on my flight, I threw together this alternative.
The only point I’m trying to drive home is that cleaning and organizing data is by far the task data scientists spend the most time on AND it’s the one they like the least. I’m fairly sure no one surveyed has used Alteryx because, if they had, the percentages would swing dramatically towards mining data and defining algorithms, which is really the important work data scientists do.
So here’s the question, how do we get Alteryx into the hands of those people that traditionally code all of their data prep? How can we enable them to do more impactful work? The answer is easy really, they need to do a 14-day trial of Alteryx and give it all they can for those 14 days. I’m confident Alteryx is the tool to solve the imbalance in their work.
June 4, 2016
Paris! It evokes images of beauty, tranquility, reverence, history. For our kids, this was a must-take holiday, so this past week we traveled to Paris for half term (and Henry’s birthday). When we take a holiday, it’s generally not relaxing. Often we need a holiday from our holiday. Paris was no exception.
Despite the terrible rains, we toured museum after museum. And I tracked it all on Foursquare, with each check-in getting logged to Google Sheets automatically via an IFTTT recipe. This allowed me the chance to test out the new Google Sheets connector in Tableau 10 as well as the improved Mapbox integration.
Tableau Public isn’t running the beta yet, so for now, this image will have to do. It was quite the trip! 26 places visited in a just five days (though this does include food places). One day our kids will appreciate this adventure they’re on.
June 3, 2016
This week my wife and I took the kids on holiday to Paris for the first time. In our heads we pictured beautiful, sunny, warm Paris full of outdoor activities, baguettes and warm croissants. But Mother Nature had other plans and dumped tremendous amounts of rain on us. We literally watched River Seine rise and by Wednesday morning, she had begun to breach the lower banks. On Friday the Louvre even closed and had to move many pieces of art from its basement.
As I arrived the airport Friday morning for my flight back to London, I downloaded historical rainfall data for Paris and put together this Tableau story. The news story have made the floods sound terrible, and they are, but I wanted to know how terrible the floods are in historical context.