Skip to content

Retail and Consumer Goods TUG

MiniTUG: Dashboard Containers and Blue vs. Green Pills with Matt Huff and Blake Feiza

36 min 6,174 words transcribed

Summary

A casual MiniTUG from the Retail and Consumer Goods Tableau User Group, built for analysts getting comfortable in Tableau. Co-lead Matt Huff opens a blank dashboard and walks through containers: deleting the phone layout, choosing a fixed size such as Letter Portrait for one-page recaps, why horizontal and vertical containers start out identical, and how Tableau's default tiled hierarchy turns messy as objects pile up. Matt's fix is to shift-drag a floating container onto the dashboard as a clean base, stretch a second container off the edge so legends and parameters land in unused grey space, set formatting once with Format Dashboard, and double-click a sheet's grip to step out through the containers it sits in to adjust padding, backgrounds and corner radius. Co-lead Blake Feiza adds three shortcuts: press the asterisk key twice to expand the whole item hierarchy, use the mouse scroll wheel over a padding box, and press T to toggle the Dashboard and Layout panes. After Matt Meier asks whether floating or tiled performs better (it scales with the number of objects, not the layout), Blake takes on one of Tableau's most misunderstood fundamentals: blue and green pills don't mean dimension and measure. Blue is discrete and green is continuous, dimensions split the view while measures aggregate, and the two ideas combine into four cases, which Blake demonstrates live with Superstore.

Listen to this TUG

Prefer to listen? This meeting is also an episode of the Tableau TUG recordings podcast.

Key points

  • Delete the phone layout on a new dashboard unless you are actually building for mobile.
  • Horizontal and vertical containers behave the same until you add a second object, which forces the direction.
  • Tableau's default tiled hierarchy gets complicated fast; start from a floating container sized to the dashboard as a clean base.
  • Stretch a container past the edge of the dashboard so legends and parameters land in the unused grey space instead of the view.
  • Format Dashboard sets backgrounds, titles and fonts for every worksheet in one place.
  • Double-click a sheet's grip to select the container it sits in, and again to step out to the parent container.
  • Shortcuts: asterisk (Shift+8) twice expands the whole item hierarchy, the scroll wheel adjusts padding, and T toggles the Dashboard and Layout panes.
  • Floating versus tiled makes no meaningful performance difference; performance scales with the number of objects.
  • Blue means discrete and green means continuous; neither color means dimension or measure.
  • Dimensions split the view and create marks; measures aggregate, and either can be discrete or continuous, giving four combinations.

Jump to a chapter (21)

  1. 0:00 Welcome to the MiniTUG
  2. 1:29 Today: containers and blue vs. green
  3. 2:01 Starting a blank dashboard: phone layout and sizes
  4. 3:24 Horizontal vs. vertical containers
  5. 5:00 Why the tiled hierarchy gets messy
  6. 7:33 A floating container as a clean base
  7. 8:58 Parking legends off the dashboard
  8. 10:04 Format Dashboard and worksheet formatting
  9. 11:41 Double-click to select containers, padding and borders
  10. 13:55 Wingdings, Unicode and Alt-code bullet points
  11. 15:55 Blake's shortcut: asterisk twice to expand the hierarchy
  12. 17:23 Scroll wheel padding and the T key
  13. 18:29 Building from the inside out, and tiled vs. floating performance
  14. 21:23 Blue vs. green: what the colors really mean
  15. 23:11 Dimensions split, measures aggregate
  16. 24:00 The four combinations on one slide
  17. 26:35 Live demo in Superstore
  18. 31:28 Key takeaway: not just binary
  19. 32:13 Right-click drag to choose an aggregation
  20. 33:38 A foundational concept, not a beginner one
  21. 35:06 Wrap-up

Transcript

0:00 Matt Huff: These, these mini TUGs, right, are, are meant to be casual, informal conversations. The other, the other piece is, is if you have, if you have anything to contribute, feel free to contribute. Um, share your thoughts, perspectives, um, anything's, anything's fair game. The other thing that we want to do is we wanted to have this as a, as a forum for, for folks to be able to look back at the YouTube videos and, and do all of that kind of stuff if they, um, you know, for, for learning and, and all those kinds of things.

0:31 So with that, um, one of the things that we, one of the thi- one of the things that Blake and I have done in the past is we've shared, um, we have shared with some of the, the folks that we work with just, like, a, a crash course on how to work in Tableau and how to do, how to just navigate Tableau at, from the perspective of an analyst.

0:54 Um, so as we think about, as we think about what that means, that means, uh, like, a lot of the, at least for me, a lot of the work that I've done in Tableau in companies that I've worked for in the past has really been more around the analytics space as opposed to the, the reporting and, and BI space. And so I use, I use Tableau to, to help build speed to insights and, and that kind of stuff.

1:21 And so that's, that's kind of the perspective I approach Tableau in. Um, i- if that, in case that's helpful context for anyone.

1:29 Um, but today what we're gonna talk about is we're gonna, we're gonna pivot a little bit. We're gonna talk about containers, and we're gonna talk about, um, blue/green and some of the, some of the pills and structures that you need to, to understand to, to better optimize your workflows as you're thinking about how to, how do I show the data that I want to show when I want to show it and, and be, be a little bit more effective with that?

1:54 Um, so what I am going to do is I am going to pull up a blank Tableau sh- dashboard, and we're gonna just start with Containers.

2:05 So just as you're working through dashboards, there's a couple of things, um, a couple of things just to kind of keep in the back of your mind. Right here, um, actually, let's just start off here at the top. Uh, phone layout. The very first thing you can do is you can just go ahead and delete that. You probably, probably won't need that. Um, uh, if you're building a mobile, mobile layout, maybe, maybe keep that, but generally speaking, 9.9 times out of 10, you're, you're not gonna need it.

2:34 Um, the other thing that you can do is you can… You've got all of these pre-filtered, um, desktop sizes. So ones that I tend to use a lot are PowerPoint as well as this, uh, where is it? Like, these, these letter portrait, um, largely because it, effectively what that is, is this is something that somebody could go print off. So if I put a report together, an analytics recap together, a one-pager outlining behavioral trends or whatever, I'll use this one a lot to kind of walk my audience through the, the one-page recap of the results.

3:09 Um, also use PowerPoint for PowerPoint, right? Uh, should be no surprise there. But today, for what we're looking at, it doesn't really matter. We're just gonna do a desktop browser.

3:21 Um, and then there's a couple of different things. So if you look over here on the left, you've got a bunch of different objects, and one of the– We've got horizontal and vertical containers. Um, and then we also have, if I come over here and click on Layout, we also just have kind of this item hier- hierarchy.

3:40 Those are gonna be important to k- keep in the back of your mind as we're working through, as we're working through this. Um, the first and foremost, as we look at Tableau, there s- as we're initially picking containers, there's literally no difference between a horizontal and a vertical container. Tableau treats them the same.

4:00 Um, so if I came out here and I… So y- I pulled over a horizontal container and drug it onto the, onto the dashboard. If I come out, come out with, um, a text and just said, "Hello, world," um, and just add that there, and then I wanted to add another text box, you see with the gray highlight, I can put that where– in either direction. I can put it horizontally, or I can per- put it vertically.

4:29 That's what I'm talking about where, um, I put in a horizontal container, but if I, if I ch- put this, um,

4:38 put this in here as a vertical, as a vertical space or as a vertical object, now all of a sudden my container is force fit onto vertical, okay? Other things that you can do is you can look at distribute contents evenly. This is, this is re- really helpful as you start to structure and, and lay out your, your dashboards.

4:59 Um, one other thing just to kind of keep in the back of your mind as you're working through this is by creating this layout, um, Tableau has this default hierarchy. And I'm not really sure what that default hierarchy does other than it is complicated for a whole bunch of unknown reasons.

5:22 Um, but as I start adding in new, new spaces and new objects, what you'll see is this, this tiled container gets to be really weird.

5:33 Um, you end up getting a lot of really complicated, complicated structures here in the tiled space, and it becomes really difficult to navigate.

5:43 Um, one of the things that you should, as you're working through a dashboard, one of the things that you should do is you should, you should try to keep your structures in relatively clean.

5:54 And so what I'm gonna do here is I'm just gonna build a couple of, a couple of quick, a couple of quick bar charts.

6:02 And, and so in this case, we're gonna look at a histogram, and we're gonna look at, uh, sales- By profit.

6:13 Okay, and then what I'm gonna do is let's look at, um, let's just create a bin and change this to Count Distinct.

6:25 And so now we've got 800 customers, um, 800 of our favorite Superstore customers, and we're just gonna change that to bin.

6:35 And,

6:40 uh, and then let's look at sales over time.

6:46 And we're gonna do this by order date, and we're gonna do just a, a simple cycle plot.

6:53 Okay.

6:56 So if I were to, if I were to go in here and just add these willy-nilly, right? And then if I come out here, and I just cha- and let's just chain, put sales on color, um, and then let's drag sales back on.

7:15 Right? I'm starting to get a whole bunch of weird structures starting to show up.

7:19 Um, if I go in here to my tile, I've got a horizontal container, I've got a tiled container, I've got a vertical container with my sales. I've got all of these weird structures that aren't really terribly clear on what's going on.

7:33 So what I'm going to do is I'm gonna clear my works- workspace, and I'm going to dra- I'm gonna shift-drag a container onto my, onto my worksheet or onto my dashboard.

7:45 And what I'm gonna do is I'm just gonna size this to be, to fit on my dashboard, put it up in the upper left-hand corner, and I'm 1000 by 800.

7:58 And cool. So now, what I've done is I've set a container up on top of my, on top of my dashboard. So if I were to go drag anything in here, I've now got a container that's going to pick that up instead of that default, um, Tableau, um, tiled hierarchy.

8:18 And so if I came out here, and I put in my total customers, I could put in my total sales, I could put in my, my total profit.

8:29 You'll see that all of, all of those charts now are showing up in my vertical container.

8:36 The other, the other fun trick that you can look at is, um, this sales container here, or this, this vertical container. And so if I click and Shift+Drag off of that, that now becomes floating.

8:49 Um, you'll notice that Tableau still has this weird structure in the background.

8:54 Um, you can, you can either remove it or just ignore it. Um, but one of the things that I'm going to do is I'm gonna take this container here, and I'm going to stretch it out over the side of my dashboard. And so you'll see that this is now extended beyond the side of my dashboard and move that over. Now, any time I pull in legends or, um, any kind of, of other, other

9:25 additional information around charts that I may be creating, uh, instead of that showing up in the background of my dashboard, that's gonna show up here on the right in the, in the gray space that isn't actually really usable space.

9:38 Um, and it's just a great way to keep your, your dashboard clean without having to, without having to worry about, um, different, different legends and, and parameters and all of that kind of stuff showing up where, where you don't necessarily want it.

9:54 Okay?

9:56 Um, other things that you can, you can do with containers is you can start to, you can start to look at, um…

10:03 So I'm, I'm a sucker for a dark background, so I'm gonna come out here, and I'm gonna change the, the worksheets to, um, to, to not have any kind of background fill color.

10:17 And then I'm gonna come up here, and I'm gonna go to Format Dashboard.

10:21 And I can change that background to be dark. I, I could change it to be pink if I really wanted to. Let's, let's do teal.

10:29 And what I can do is I can change, I can go out here, and I can change the, the formatting of different title, worksheet titles and all of that kind of stuff all in one place, so I don't have to go do that in individual, um, individual worksheets. Other places that you may wanna, may wanna explore is, like, if you format a worksheet, um, I can change all of the formatting to, to white.

10:55 And if I wanted this to be, um, if I wanted this to be Wingdings, I could do that.

11:00 Um, I wouldn't recommend that in a production dashboard, but you could do it.

11:04 Um, and then the other thing that you can do is you can come out here and, and if you, if you go, come down here, you can change all of the different, all of the different structures. You can apply a color palette to your dashboard. So maybe I want Lightning Water for whatever reason. I can change all of my axis ticks. Um, this one is one that I, I will do fairly regularly is I'll drop the, I'll drop those background images.

11:30 Um, other things that you can do with a container, so is I'm going to… Or I guess this isn't, isn't so much a container, um, but what I can do is I can come out here, and I can click do- so if you look at this, I've got a worksheet highlighted. If I double-click this worksheet, you'll notice it's gray right now. If I double-click that, I get a blue background, um, for that, for that, um, whe- when I click on that, double-click on that, that little, I don't know what you would call this.

12:04 It's not really a hamburger, but, um, we'll call it a hamburger for now. If I double-click on that, what, what ends up happening is I end up selecting the, the container that that worksheet sits in. And so just as an example, if I came out here and I clicked and brought in another horizontal container, and I moved this sales data here, and then I moved that profit data there, if I double-click, you'll now see that I'm selecting that container that they're all contained in.

12:36 I can distribute contents evenly. And then if I double-click this again, I then go out to that master container, and I can change this to, um, I can start to then change whatever padding I want. And so I'm gonna change the background to be maybe a, a dark gray, and I can then change my inner padding and, and we can start to see, we can start to shift some of those different, different pieces around.

13:05 Um, if I do outer padding, I will get the, I'll get that, that background, and then I can change, come in here and I can change the, um, the fillet around the edges. And so now you can see that I've got a, I've got a dashboard, um, two-tone dashboard with, um, different formatting all set at that container level. I could then come out here to this, to this smaller container, and I could do, I could do something similar, right?

13:34 Um, maybe I want the background here to be a light gray, and I want to change the padding to be, uh, 20, just as an example. And then I want to have another, another fillet around the edge.

13:49 Um, I could, I could do all of those kinds of things, um, with, with containers.

13:55 Cool.

13:56 Any questions on any of that

14:00 other than why I use Wingdings?

14:05 Blake Feiza: You know, there, there is actually times where I have used Wingdings legitimately in a dashboard because it has certain, uh, characters that you can use in-

14:17 Matt Huff: Mm-hmm

14:17 Blake Feiza: … in place of, like, arrow- arrows for ups and downs.

14:22 Matt Huff: Mm-hmm.

14:23 Blake Feiza: The trick was trying to remember to, uh, set your label in your text box to be Wingdings, and at the appropriate Wingdings. I think there's, like, three different Wingdings fonts that Microsoft generally has installed on their products, and I assume macOS users have the same. That's a question mark. You know, I don't know.

14:48 Matt Huff: Yeah, I would, I would assume.

14:50 Yeah, I would assume there's, there's a, a use case for some of them. I, I have had to use Unicode characters. Um, and-

14:57 Blake Feiza: Mm-hmm

14:57 Matt Huff: … a lot of times what you can do is you can, you can look those up.

15:00 On, on a PC, if you've got a number, if you've got a number pad, one of the things that I really like to do is if you, if you hold down Alt and then sl- and then pick some of the numbers, um, you can auto-generate bullet points and, um, a- and whatnot off of that number pad.

15:23 So there's some, some good stuff to, to look at there. This, and, and so for me, again, because I'm using, because I'm typically using Tableau as an analytics delivery mechanism as opposed to, like, a, a BI-type reporting platform, I, I end up using things like bullet points and a lot of that kind of stuff pretty regularly in my dashboards.

15:47 And so those are, those are just some, some things that I tend to, tend to use.

15:53 But cool.

15:55 Blake Feiza: Sweet.

15:56 And I have, I have three quick tips too on containers, Matt, that I can share just to-

16:00 Matt Huff: Yeah, go for it

16:00 Blake Feiza: … just to jump in and maybe provide more value. We'll see what you guys think. Bear with me. Uh, the tips won't be that, uh, you don't need to see anything super close. Uh, but one thing that I wanted to call out is I've replicated some of the, the disaster of Tiled over here by just dragging a bunch of sheets on and text elements.

16:20 And if I go in here inside of Tiled, right, just like Matt showed, there's a whole bunch of mess in here.

16:26 Uh, but one thing that can be a little bit annoying is opening this up for the first time and fanning it all out. So a shortcut, and I don't know why this is a shortcut, is if you do Shift +8 or, like, the asterisk key two times. The first time I do it, you see it opens one layer below Tiled. The second time I do it, it actually fans out all of the objects inside of this hierarchy.

16:48 So this can be super useful. I learned this from Jeff Shaffer at a Tableau Conference. Uh, if you want to go through and, you know, maybe find different elements, delete things, uh, it's much quicker than clicking all of these little carets. So that's a fun little tip.

17:03 Matt Huff: Uh, does double-click do anything? 'Cause I think double-click opens the first layer. I don't know if it does it again or just closes the-

17:08 Blake Feiza: Yeah.

17:08 Matt Huff: Yeah.

17:09 Blake Feiza: So double-click will do the first layer but not that second full fan out. You have to still go through and double-click them one at a time. But for some reason, the astr- asterisk key twice opens everything up.

17:21 So that's a fun one.

17:23 Um, another one that can be quite nice is if you've ever tinkered around with the padding but you don't know quite what you want, so maybe, you know, you go from 4 to 15 or I gotta backspace and do 20, uh, one thing that can be really nice is if you actually have a mouse, uh, and you use your scroll wheel while you're hovering over one of these, uh, you can scroll wheel up and down with the padding, uh, to get that pixel perfect alignment with different things.

17:48 Uh, so that's a fun one. Uh, and the other one is if you want to switch between the dashboard and layout panes up here, it can be inconvenient to go all the way up to this top corner.

17:58 Uh, so if you just hit the T key on your keyboard, uh, that toggles between these two options as well.

18:03 So couple little tips. Um, Matt covered most everything that I would have said too around avoiding Tiled.

18:10 Uh, generally, I, I can't think of a time that I wouldn't start with a floating container and then just kind of treat that as my base tile.

18:18 Um, so- A couple different tips. Um, yeah, take 'em, take 'em or leave 'em. None of those are going to change the game by any means, but it might speed up your workflow a little bit, so, um.

18:30 Matt Meier: I'll tell you one of my tricks that I've done when it comes to building out a dashboard view is I tend to think from the inside out, and I find that if you start with, like, the bottom layer of a dashboard hierarchy and then you start building that out, granted, I use everything as a floating, so that's kind of where I'm operating from, but if you start from the inside and work your way out, it tends to work a little bit better.

18:57 Um, that's just me. Might be helpful tip for everybody else. But I, I did have this question I wanted to ask. Is there any meaningful difference in performance between a tile versus floating?

19:09 Blake Feiza: As far as performance goes, I wouldn't say there's anything dramatically different that you'd see. It, it… I think it more so scales with the number of objects that exist on the dashboard. Uh, but whether or not those objects are floating or tiled, as far as I know, it won't make any meaningful difference in performance.

19:27 Uh, there's probably a, a super nio- nuanced answer, uh, that is one is slightly better than the other, but I, I couldn't tell you which one.

19:38 Matt Huff: Yeah. I don't, I don't have a, I don't really have a reason for, for one versus the other. I think the, the approach that, the approach that, that I walk through, I, sometimes I use it, sometimes, depending on, depending on how rigorous I need to be, a, a lot, like, I may, I may just suffer through the, the tile mess that Tableau, um, gives me.

20:03 Um, I haven't, I haven't really needed to use floating outside of…

20:14 So, like, in my, in my Makeover Monday this week, I, I needed to have some labels put in different places, and so I used s- floating for things like that. Otherwise, I can typically accomplish most of what I need with, um, tile.

20:29 Blake Feiza: Nice. Perfect. Perfect. And one of the things too that drives me crazy with tile, I was just looking to see if I could edit the height of this sheet here by clicking this caret, uh, which is not showing up on the stream.

20:41 Okay. Well, if I click this caret right here, typically there's an option that shows up for fit width or fit height or edit width, edit height.

20:50 Uh, but because we're in tiled, like, there's something going on with this sheet, um, where, you know, because it's in this tiled object, I don't actually have that option that appears. I wish I could show it on the shared screen, but it's not appearing. Um, but you get the idea. Uh, starting with, like, that base floating container makes it way more consistent for what, uh, you know, what those settings look like as you're building out a dashboard.

21:17 Sweet. Uh, with that, I will… I can't flip over while I'm sharing like this.

21:22 So one of the most important beginner tips that I learned way later than I should have with Tableau is what the blue and green colors actually mean inside of Tableau.

21:36 So as a beginner, pretty much everybody assumes that the blue fields go up to the top of the view in Tableau on the left, and down below you see the green colored fields. Uh, and it basically, by default, people assume that if you flip to the next slide here, Matt, um, blue equals a dimension up top, and green is all of your measures down below.

21:57 I am here to tell you that is not the case at all, uh, and it actually is a, a super nuanced kind of fundamental that can really change the way that you build visualizations inside of Tableau.

22:07 Uh, so in reality, blue represents a discrete field, and green tells you if something is continuous.

22:15 So in fact, you have kind of a, a hierarchy or, like, a quadrant chart of, you know, continuous dimensions versus discrete dimensions and the continuous measures versus discrete measures. So Matt, if you flip forward a slide, uh, I have tried to create a visual that might help explain that.

22:35 So when we think about what blue and green means inside of Tableau, I'll jump into a dashboard and, and we can kinda drag some things out to see this more visually.

22:45 Um, discrete i- is going to be blue. So that's essentially anything that creates some sort of a new header inside. So if you think maybe dragging region onto the view, uh, by default in Superstore, it creates those headers or, or the groups of regions in a list.

23:01 Uh, and then anything that is green is going to create you an axis, so something continuous that gives you, you know, that, that axis value.

23:12 On the other side, uh, these dimensions and these measures that people often confuse with being mapped to colors, uh, what a dimension does is this splits the view. Anytime that you add a new dimension in, whether it be continuous or discrete, um…

23:28 Yeah. Whether it's continuous or discrete, when you drag a new dimension into the view, uh, it's going to split up your view or create new marks inside of it.

23:37 Measures, on the other hand, always aggregate your data. So if you think about a sum or a count distinct, anything like that, those values can actually be both blue and green, uh, but the way that they aggregate your data up is what makes it a measure. Uh, it's that aggregation piece.

23:55 So if you flip forward one more slide here, uh, this is es- essentially what we can demo inside of Tableau, but visually, this is kinda what we're talking about. In that top left quadrant, uh, you'll see on the left there's a customer name row. So that's if we, you know, go and drag customer name up to our rows, we'll see that list of customer names all fan out.

24:18 You can do something very similar with perhaps, like, the sales field or the discount field. Uh, and instead of getting an aggregated value- Uh, a dimension is going to split it up. So in this case, in the screenshot, it shows you every individual value of sale that appears inside of your dataset, the same way that it would for every individual customer name inside of your dataset.

24:41 Uh, moving over to the top right side, uh, we have now a continuous dimension.

24:47 So again, we're not aggregating. Uh, it's a dimension. It's not aggregated.

24:52 So all of our sales values on that left chart are showing up as an axis. We can see that the Y-axis goes up and plots sales.

25:00 But instead of having one mark with like, say, a sum of sales, we have every individual mark representing all of those sales values that we see, kind of like that table to the left of it.

25:11 Uh, the same sort of thing can happen if you maybe plot like sales versus quantity or something, we can get a scatter plot where all of those individual marks are fanned out because there's no sum or average or min or max. Uh, we're actually plotting unique values onto, uh, onto the view.

25:32 Um, and then down on the, the measures side of things, this is where we get into aggregations, right? So a discrete measure would look like a BAN. So if we want a big sum of sales number to show up on our text shelf, uh, we don't want that to be like an aggregate that gives us, um– or sorry, we don't want that to be continuous where we get like an axis showing up.

25:52 We would just want that to be a discrete value giving us our sum of sales or, or profit or count distinct of Order ID in that bottom example. Anything that is aggregated, but we don't want an axis becomes a blue discrete measure.

26:08 Uh, and then down in that bottom le– right is kind of that more common case, right? Where we have our measure, and we want it to show up on an axis. Uh, so we've got, you know, maybe sum of sales or count distinct of Order ID, and we get that axis. So that's a whole lot of talking about a slide. Hopefully, this is something that can be useful if, if screenshotted or, or kept in, in a back pocket.

26:29 But I can demo what some of this looks like inside a Tableau, and that might be a bit more useful.

26:36 So let me steal this back.

26:39 I'm gonna try and switch monitors, actually. Let's see if that fixes the problem with the blurriness.

26:45 Oh, it looks even worse. Oh my gosh. Why is it doing that?

26:51 Okay.

26:51 Well, the, the downside is when it's blurry like that, we can't actually see, like right click actions. So like if I right click here, I have like a, a list of things that are showing up, and they aren't visible.

27:05 Um, but you're gonna have to just trust me, uh, when we go through some of these examples because that's just kind of the, the nature of Bevy right now.

27:14 So starting with that, that top left quadrant, I wonder if there's– I can't really get both of those in. I would put it side by side, but it won't work here.

27:24 Um, if we were to take something like our category, right? This is up top. It's above this line here. So this tells us it's a dimension.

27:33 Uh, and because it's blue, we know that it's discrete. So discrete dimension here of category. If I drag this up to rows, here's what we see, right? We see that dimension splitting out the view and creating three distinct, uh, rows in this little table here.

27:50 The same sort of thing would happen if we took sales, and let's say we dragged sale– we drag sales up to the top above this line. We can switch it from a measure to a dimension.

28:03 And now by default, Tableau's actually turned it blue for us. So now we have a discrete dimension up in the top for sales.

28:11 And if I drag that up to rows, we again see a table where every individual value, and these are formatted to look a little bit more, uh, aggregated, if I make this just like a standard number.

28:23 Uh, we can see all of the individual sales values in the dataset show up separated out here.

28:30 Um, so this is again that discrete dimension. Uh, now we can keep this blue, right, and keep it discrete, but turn it into an aggregation. If we right click, uh, and trust me, I've right clicked it, uh, and we select measure and maybe make this a sum.

28:47 Now, by default, Tableau turns this green, and it makes you think that maybe, you know, green is what matches up with a measure here. But in reality, it's not. Uh, I'm gonna switch this from green, uh, to blue by changing it to discrete. And now we see that BAN number, right? We have a big value showing up here for our sum of sales.

29:07 Uh, it's aggregated, uh, and it is a discrete measure that we have now inside of our view.

29:14 Uh, if I hit the undo button, again, we have now an axis showing up. So this is kind of your standard, uh, if I create a new, you know, profit and I drag that in, Tableau gives us that axis by default, that sum of profit, um, which is great.

29:31 If we were to go into the sum of profit and change this to a dimension, what we see now is, again, it's, it's a dimension, not a measure, so there's no aggregation applied to profit. And instead, we see a mark for every individual value of profit inside of this dataset.

29:49 We could do something very similar if we took, I don't know, maybe quantity and drag that up to columns. And again, we remove the aggregation of sum by changing it to a dimension.

30:00 And now we have a scatter plot where we have individual values of quantity, individual values of profit, all mapped out against one another.

30:08 Uh, and you'll notice that we don't have to put anything like customer name onto our detail shelf here because these aren't being aggregated. We're already seeing in the dataset all of the unique combinations of quantity and profit on a scatter plot.

30:22 So it's slightly different than maybe dragging, you know, quantity up and profit up.

30:29 Uh, when these are our measures, we have that sum aggregation applied here.

30:33 And we get one mark initially because there's nothing splitting up the view.

30:38 Uh, but if I were to grab maybe like customer name and add it to detail, now we have one mark for every customer name inside of that data set, so that's effectively fanning things out.

30:50 You would see something similar here if I added customer name to the detail, I would fan out the data slightly. You see a little bit of movement there.

30:58 Um, but you can see the marks in the bottom corner here go up when I added, um, customer name onto the view. We go from 7,600 marks up to 10,100. So it is, it is breaking that view out a little bit further, but without it, we're already getting all of the unique combinations of quantity and profit, uh, in this scatter plot.

31:21 So I think that's all of the different examples.

31:26 Um, skimming at the slide.

31:29 Yeah, essentially, the important thing to know here is that blue and green does not mean dimension and measure. Uh, we don't need to have only blue things up top here. If I change sales to be continuous, you can see that we have now a green value up on the top.

31:44 Uh, and we could do the same thing down here with maybe quantity if we wanted this to be discrete. Now we've got a blue field down on the bottom.

31:51 Uh, essentially, there's, there's four different things here going on, not just, it's not just binary.

31:57 So that's the key takeaway here.

32:01 Um, happy to, happy to expand on anything as well or if any, if anyone else wants to chime in with thoughts for general education, that is, is always welcome, too.

32:13 Matt Huff: Any thoughts, comments, questions?

32:17 Right click

32:19 Matt Meier: I just dropped that in the chat, but that is a helpful tip that I don't remember how I discovered. I might have done it by accident one time, but if you right click and drag to there, and it obviously it's not showing up for you, Blake, but you'll get a context menu that allows you to select, uh, different aggregations. So if it's like a continuous pill, and you do that, it will allow, um, you to select like sum, average.

32:47 Yep, there it is. Or you can do the non-aggregated view for profit at the very top.

32:53 It's very handy when you're trying to build something quickly, rather than just doing like a simple double click and have it auto add.

33:02 Matt Huff: Well-

33:03 Matt Meier: Yeah

33:03 Matt Huff: … and just to, just to call that out, right, is that whole, that whole profit at the very top, right? That's the, that's that discrete, discrete piece that Blake was talking about earlier, so.

33:19 Blake Feiza: Yeah, specifically the dimension piece, because it would still be, uh, continuous, but it would be a continuous dimension that this would pull in.

33:29 So that gives us the, the continuous piece gives us that axis, uh, but the dimension breaks up the view into a bunch of different marks.

33:38 Matt Huff: Yeah. And, and Paul, I wholeheartedly agree. Like, I don't, I don't know that this is necessarily beginner, right? But it is, it is one of those things where as you're, as you're running and as you're building out visualizations and whatnot, like you surprisingly run into this more often than not.

33:59 And so it's really helpful to have at, at the very least just to be exposed to it, to know, hey, I remember hearing this one time, maybe, maybe I should go back and, and look at this again, so.

34:14 But yeah, I, I wholeheartedly agree. This is like, this is not what I would consider like super, super basic by any stretch of the imagination.

34:25 Blake Feiza: Definitely. Yeah, I, I would consider this a, a foundational concept that might not be super simple to grasp, but there's a lot of, of Tableau principle that is built on top of it.

34:35 So I think having a firm understanding of, of what's going on underneath the hood here, uh, can make learning, you know, much easier once you kinda get that, that baked into your brain and into a rhythm there.

34:48 Well-

34:48 Matt Huff: Any other thoughts, comments, questions?

34:52 I think, I don't have anyth- Blake, anything you wanted to, anything else you wanted to add?

34:58 Blake Feiza: No, I think that, that pretty much covers the, the curriculum we were hoping to get through today, so.

35:04 Matt Huff: Okay. Cool.

35:05 Blake Feiza: I'm good here.

35:06 Matt Huff: Awesome. Well, then let's, let's go ahead and wrap this up, 'cause we don't need to, we don't need to take everybody's full hour. Again, like, like we were mentioning at the beginning of the hour, um, these are, these are fairly casual, right? And so if we finish early, we finish early. If we, not that we wanna run long, but, um, if we've got questions, we'll, we'll go ahead and answer them.

35:27 But, um, appreciate everybody, everybody showing up and, and participating. Hopefully it was, hopefully it was value added.

35:35 Um, and if there's something you'd like for us to cover in a future session, let us know.

35:40 So, all right.

35:42 Thanks everyone.

35:43 All right.

35:44 Blake Feiza: Thanks, all.

Shorts for this TUG

Leave a comment

Your email address will not be published. Required fields are marked *