Break time! Nothing fancy for me today. They’re so many choices. I could be here for weeks trying to decide on the best one; which gets me nowhere. I just grabbed a cup from the water cooler.
- Get Started with Tableau Desktop
- Step 1: Connect to your data
- Step 2: Drag and drop to take a first look
- Step 3: Focus your results
- Step 4: Explore your data geographically
- Step 5: Drill down into the details
- Step 6: Build a dashboard to show your insights
- Step 7: Build a story to present
- Step 8: Share your findings
- Learning Library
So my first real dive will be with this tutorial. It’s cited in several places. In fact it’s mentioned just 1 click away from the Community’s landing page. You know, where it says “New to the Community?”
First paragraph and I’m in love – a real world working example. Hello World has it place. But this tutorial is giving me a real situation, describing my goals, and telling me what to expect. A story I can follow and empathize with.
Tableau Desktop comes installed with sample data? Bonus. I don’t have to go searching online for the example’s download site; or try to figure out which version of the example to use. I’m looking at you Adventure Works.
It’s also reassuring me that Tableau doesn’t alter my data. Knowing this empowers developers to explore and experiment since they don’t have to worry about accidentally altering the source data.
In several places the tutorial gives me expandable sections called “Learn more”. It suggested not to dig into these on my first time through, but I can’t help myself. Personally, I’d recommend it, because it explains the UX/UI, terminology, and why Tableau arranges things the way it does.
There’s even animated gifs to further demonstrate the UI and review the instructions I was supposed to follow.
Is the rest of tableau.com like this? If so, this is going to be great.
And then I came to a screeching halt. And it was my fault. But I didn’t figure that out till later.
As I followed along, it was obvious my worksheet didn’t look like it should. The tutorial was trying to demonstrate how Tableau revealed a loss in profit; but the data given had no such information.
First, it turns out the data from the online tutorial didn’t match that which was installed with my version of Tableau Desktop – 9.3.5 (9300.16.0726.1843) 32-bit.
The tutorial has orders from 2015, but the latest order I had was on 12/31/2014. The tutorial didn’t supply a copy of the data I could download; so I searched both tableau.com and my PC. There were a few, but none of them seemed to match up exactly.
Frustrated, I started over; and low and behold, it worked. I still didn’t have 2015 sales data, but my workbook was acting very much like the tutorial.
Root cause? While playing around in the “Data Source” I was double clicking tables trying to display the information and take a look at the raw data. Sadly, when I double clicked on Orders; Tableau joined the Orders table to itself. Ugh, what’s in this water?
Don’t blame the water Marshall. It was your own fault. Next time you experience strange behavior like that, try looking at the data first. You know: is the correct data being used, is it joined correctly.
I think I did a spit-take at how easy it is to create a map.
Everything looks great. Mostly. Profit’s color pallet is set to “Automatic”; but it looks like my worksheet is using “Red-Green Diverging” for its “Automatic” setting instead of the “Orange-Blue Diverging” as the tutorial said it would. No problem; I’ll just edit the colors from Automatic to “Orange-Blue Diverging”. There. Now my map looks like the one from the tutorial.
Tutorial's worksheet with Automatic color palette
My worksheet with Automatic color palette
My worksheet now using Orange-Blue Diverging color palette
Oh Marshall. Didn’t you learn your lesson before about going off script?
Here the tutorial is showing me how to use a worksheet to investigate data. But, it didn’t point out that the worksheet as designed, couldn’t be reused if the data gets updated. Harrumph # 1.
At the end of this step, the tutorial is showing how FL, SC, and TN can go from negative profits to positive by removing binders, machines, and tables. But, instead of having all southern states showing a profit, my map looks like there are are now 8 states with negative profits.
Wait. It's the coloring. The tutorial’s map shows orange for negative profits. Mine is just using orange as a lower bounds. If I change the profit’s color to “Automatic” it’s no longer misleading. Harrumph #2.
My worksheet using Orange-Blue Diverging color palette
My worksheet using Automatic color palette
I’ll drain the water cooler if I keep going down rabbit holes. I’ll cover these 2 issues later. I’ll have to think what beverages would be appropriate. Yes, yes; I’ll update this blog and insert links later.
Anyway, back on task.
In this step filters on one worksheet are now applied to all worksheets. Interesting, but confusing too. I’m expecting worksheets to be isolated and not affect others. It’s a powerful feature; I’ll just investigate it later.
OK, seriously!? All the data in this tutorial includes 2015 sales; and mine doesn’t. Not perfect, but I was dealing with it. Then at the end of this step, I was presented with an embedded version of the story just created. And it only has data up to 2014.
Not only that, its “Automatic” colors are the same as mine – “Red-Green Diverging”. Harrumph #.
Calm down Marshall. You keep that up and you’ll have to rename your blog to Harrumph Central.
I love that this tutorial acts as a working example; which of course means publishing your work.
Oh, that doesn’t look quite right. Er; minor adjustment, save, delete, and republish.
So, make sure to select the first sheet before publishing. Whichever sheet is in view when you publish, will be the starting point used on the server.
I almost skipped the learning library. Big mistake. For all the features covered in the tutorial, it gives links to the applicable page within Tableau’s Online Help. There’s also some additional links to other Tableau resources.
Wow, look at the time. Long post, I know. Till next time.