Re-Upload - Building Basic Charts in Tableau Next Part 3 [Reposted]

0:02Alright, we're back.

0:03We are going to build some charts today in Tableau Next.

0:06We've been sort of working through the experience so far.

0:09We're here on what I now call the landing page.

0:12I've been calling it a workspace, it's not, it's the landing page.

0:15When we click on one of these orange icons or we go to a workspace, this is the workspace view that we should be looking at

0:22Table next persists the tabs across sessions.

0:25So I actually had these two already open and it seems to have remembered them, which is quite good.

0:29So we can essentially carry off from where we left off.

0:33Now yesterday I looked at the semantic modeling.

0:36And I k I hit a couple of snags and actually a few people brought up some really good tips.

0:40First of all, I approached the challenge a bit like a Tableau user trying to do something.

0:45in the Salesforce ecosystem, which is kind of the wrong way to really frame this and look at this.

0:49I should really be looking at the question, um, how is it what's experience like as a Salesforce user trying to do work

0:55in the Salesforce platform.

0:57That's really the sort of the basis I should be using for this.

0:59So um whilst I was drawing parallels to Tableau and I think it's logical to do that because a lot of people want to know more about the product

1:06Going forward, that shouldn't be the basis.

1:09Now, I've said before it is buggy.

1:11I did literally nothing and it it spawned an error.

1:14So um that is absolutely fine.

1:16And I guess we'll just refresh this.

1:17But that's to be expected.

1:18This is still in beta, it's not been launched as a

1:21V1 or a polished product just yet, so bear that in mind.

1:25So to build charts, I thought about doing this.

1:29So if I just go to Tableau Charts

1:31Tableau have this guide to charts.

1:34This is sort of their canonical guide in here.

1:36Like here are basic charts you should be able to do.

1:38Pie jar, bar jar, line jar, scatterplot, histogram, bullet

1:41packed box and whisker tree maps.

1:44So I want to create these inside of Tableau next just to see what the experience is like, see how easy it is to do some of these things.

1:50And they're a nice sort of comparison point because you will know how to do those in Tableau

1:54I've got a great series showing you how to build basic charts in two minutes.

1:57So what we're going to do is go back into our workspace.

2:02Let's just go back in here.

2:03And we're actually gonna uh upload a new data set.

2:06I'm just gonna go through that experience again because I think it's quite a valuable thing to do.

2:10But this time around

2:11I'm going to use Superstore because we all love Superstore.

2:14Okay, so let's go ahead and add um let's add new data.

2:19Let's just go ahead and do this.

2:20And for this one, I've actually got a an Excel file on my desktop

2:24and go on then let's grab it in and drop it in and uh this is superstore it should have three tabs orders uh returns and people so we're just gonna go ahead and bring all of those in

2:36Uh these are going to be DLOs, so let's just go ahead and create those.

2:41I'm doing this a tableau way.

2:42Um you do this in data cloud strictly speaking.

2:45And actually if you have a Salesforce um

2:48company or org this is actually a very sort of good way of doing it.

2:52So let's wait for this to finish.

2:54Hopefully we get no bugs.

2:55We've got two for three.

2:57Let's make this three for three.

2:58Yesterday we had bugs um I couldn't get through it without

3:02Finishing the experience.

3:04This way for this.

3:05Orders is the bigger table to be fair.

3:07All three have worked superb.

3:09Let's go ahead and create the semantic model.

3:11So we'll go ahead and click on that

3:13And we'll go ahead and create.

3:15So I've simply just literally click that having turned them into data sources.

3:20So now I'm going to the next step in the hierarchy.

3:23Those three should already be there.

3:25Now my people and orders, um, it's kind of hard here because all these are named um in a weird way.

3:32So it really should like

3:33Uh I wish it I wish it just named them after the tabs because of course everything's gonna have the same file name because they're coming from the same file.

3:41So it really should name these after the tabs just to keep the the names visible.

3:44It's a small tweak, but really

3:45Like I can't read what this says.

3:47So if I go ahead and click on this, I have to hope that this is the um people table.

3:51It is when I click on that I do get the preview over here, so I'm able to do that.

3:55And so what I can say here is that I think

4:00There's like a manager column, is that right?

4:02Um is that what it is?

4:05I forget I forget who the um what the what the column relates to, so

4:11segment blah blah blah what I should really do is look at the people table and it should tell me what that should be it's region so it's region apologies so I should say in here uh region

4:25equals region I don't like that it auto selects its own join field as a default that's that makes no sense to me um

4:35Yeah, that does that makes no sense.

4:37So go ahead.

4:38I'm always forgetting to hit apply on this.

4:40Um it should just be automatic.

4:42Everything else is kind of automatic, so it kind of should just be like that.

4:45And then with the sample returns, I think it's just an order ID.

4:49So let's go ahead and select that and we can go look at the returns table and you can see

4:53Auder ID is the um uh common field, so we can just go ahead and do this.

4:59So this is what I mean.

5:00It's uh it it chooses the data source object

5:02Which is a column it's created when it's done the import, which makes no sense to me.

5:06So um yeah, we should really just be matching to fields in the table that I have not created as a system in my opinion, but that's my opinion.

5:15So anyway, let's hit add.

5:17hit apply okay so there we have our data model we'll call this sample store sample super store data model

5:26Um this is very much like what you do in the connection Mendev tableau, except for you should be building the data model.

5:33If not, go watch my video on the data model on my other channel

5:38And yeah, we can test the model, but I know this works, I'm not gonna bother with that with that.

5:42So we're gonna go ahead and do the first thing, which is to create a visualization.

5:46So by the way, as we're skipping these sort of

5:49um steps in in the workflow it is actually updating this so you can see now I have two data models I also have my orders and returns and so it's actually adding to these data objects which is important.

6:01So

6:02Now we're in a good place to actually start using this to build visualization.

6:08So I'm going to close the bookshop semantic model

6:11I'm gonna close this file because we've done the import and I'm gonna close the um data model because we've built it and I don't think there's anything else we need to do here.

6:20Actually I refrained because we might want to build metrics, so we'll come back to that in a short while because there actually is a good thing to do there.

6:27Maybe that will be a separate video.

6:29Let's just build charts for now and then we'll do that in a separate video over the weekend.

6:33So

6:34The first chart we're going to build is going to be a pie chart.

6:36So let's go ahead and go to our chart types and we'll see here we have circle, we have donut.

6:41There's no pie charts apparently on this platform

6:43form at least not yeah it's not one of the supported chart types but you can see this doesn't have the full range of chart types that you have in Tableau it has some of the more basic ones so um yeah people are gonna ask for pie charts just

6:55just out of default and the fact that there isn't one is sort of strange.

6:59Don't know why they've gone for donuts first.

7:01So let's just like select donuts and we'll just do the classic region on colour

7:06And we'll do sales on angle and that should give us a pie chart.

7:11Okay.

7:12And um what I typically do in Tableau is I put this on detail, sales on detail, then I go and do percentage of total calculation

7:20Uh there is no percent of total calculation.

7:23Okay?

7:23There's running total, difference, percent of difference, rank moving, but no percent of total difference.

7:28Am I missing something?

7:30No.

7:30No, there's no percent of total difference.

7:33Okay.

7:34So um at this point I'm like, do I do I have to create an LOD here to get this to work right?

7:41I mean that does make sense to me, but uh alright, let's try and

7:44Create a calculated field and let's see what we can do with this.

7:48So the calculated field interface is a little bit like disorienting because it's got all the same things but

7:54It's not quite the same.

7:55So hey, let's see if we can ask Einstein to create it for us.

7:59So um create a

8:02percent of total um calculation and um to uh

8:14Figure out proportion by region.

8:20Uh create a got a space there.

8:23Draft resign.

8:24Let's see what it does

8:25I can't even spell create.

8:26My bad.

8:28So sample steepest or cells divided by the sum of stepest or cells.

8:32Okay, so it's written a row level.

8:34Um

8:38Hmm, I mean this is not what I would do.

8:41It's not what I would that's not how I'd do it, but let's see.

8:45So the other thing is using a it's using it's just

8:49It's just so different.

8:50Um again, it's maybe not for me, but it is very, very different.

8:55So if can I go full screen on this?

8:58Yes, I can.

8:59So I can get rid of all the other stuff.

9:00Okay, so

9:02This calculation determines portion of cells contributed by each region relative to the tesour cells across all regions.

9:07Okay

9:09the average that doesn't seem right um that really does not seem like up as good so you can build in like sentimentality into this uh data role continuous yeah

9:21decimal places we can just do uh zero it's percentages so we don't want any and yeah you can kind of set that up so let's just let's just say okay maybe this does the right thing.

9:31It's not how I would re would have written it but where did it go even?

9:34Um

9:35What did I even call that calculation?

9:37I don't even know.

9:38Um calculator fields one.

9:41That's such a classic thing I'd do in Tableau.

9:44So um okay, can I

9:47Can I can I can I bring it in?

9:49Ah.

9:50So I created a calculated field.

9:54Um why can't I see it?

9:58What have I done wrong?

10:01Oh, it's right here.

10:04That's such a weird UX.

10:06I thought this was the calculated field.

10:08Actually it's this one sells

10:09So, okay, so I've just figured out something else out, and this was not obvious to me at all before.

10:15These are actually headers.

10:17Um it would be nice if they had a little UX arrow like on the right hand side with a down so when you click on it you actually have some sort of concept of hierarchy.

10:26Um I get that flat design is like something that's big but it is really not um helpful um if I just expand these and

10:34you just end up scrolling through them.

10:36Like it's really not helpful that it's not obvious that that is actually a hierarchy.

10:40So uh nevertheless we figured it out and uh there's a bug here.

10:44I don't know if you can see this and now I'm I'm trying to um

10:48I'm trying to expand this and it just won't expand.

10:52Okay, that's that's frustrating.

10:54Um we'll skip that for a bit.

10:56Let's go let's go to the fit entire view.

11:00Um

11:00And let's put sales on label to get the total sales.

11:05There we go.

11:06And let's see if now that's cleared my bug.

11:09It has not.

11:11So at this point, let's call this um pie chart.

11:16It's not a pie chart, but let's just call it that.

11:18Hit save.

11:19And then once I've hit saved, I'm gonna refresh the page and there you go, the proportions come back.

11:25So let's also put this on label

11:27And um it says 1% for everything, which is obviously not correct.

11:31So that was not how it's right the calculation.

11:34It was very confident.

11:34I knew that wasn't right, it didn't look right, but I tried it anyway.

11:38Um

11:39Ah, can I be bothered to write an LED at this point?

11:42This this should be this should be so simple.

11:44So um can I type in the term fixed?

11:48I can.

11:49fixed and then I can go ahead and type in fixed at the what I what I uh what I have to do here um um

12:05Do I have data source objects?

12:07Yeah, fix it the data source object level.

12:09That will do.

12:09That will absolutely do.

12:11Give me the sum of

12:16Sales.

12:20Sales.

12:22Uh yeah

12:25It's such a weird way of writing f um formulas.

12:30Gotta not gonna lie.

12:31So fix this uh the

12:34data source object level give me the sum of cells this will be my uh total cells for uh dataset and then

12:47What I need to do is go ahead and hit OK.

12:53Your attention please.

12:54Creation of semantic entity of cells at CAFAL because formula validation failure.

12:59User aggregate cannot be used in LOD

13:01Um user ag cannot be used in LOD.

13:08User aggregation cannot be used in LOD calcs

13:14But there is no user aggregation.

13:17Does it mean this?

13:19Oh, I didn't set that.

13:24Well well I didn't set that, so hold on.

13:26It's some fine

13:28How about that?

13:29Anything now?

13:30Um let's hit validate, see if it works.

13:34Okay, the validation has cleared.

13:36Why is it said to user aggregation?

13:39I have no idea

13:41Okay, that's very bizarre.

13:43So let's put that on label to see what we get.

13:47We should get the same number on every column.

13:49Great.

13:50So then what we can do is we can create a calculated fill that says uh sales, actually sum of sales to be strictly accurate.

13:58um and its sum of sales.

14:02So it's weird because you type it in and then it prefills a thing.

14:04Um divided by

14:06And then you want to go total sales data set, which is actually in the calculation.

14:11So and this user aggregation thing, I don't know what this is about.

14:15Select how to aggregate.

14:17This field defaults a user ag and can't be changed.

14:21Wait what?

14:23Okay.

14:23I mean I don't know if user ag means that I've dictated the user aggregation here, so it can't possibly tell me what it is.

14:31And then this can be set to percentage gone.

14:35Gone in.

14:36Validate that works.

14:37Oops, no.

14:38API name is required and cannot be empty.

14:43Wait, what?

14:44Why do I have to do all of this work?

14:46Um I just want to create a percentage of total.

14:51Please, please.

14:53percent uh no okay fine fine let me type then okay um percent of total it's fine no it won't let you um

15:05Do anything about that calc.

15:07Uh validate.

15:10Level error.

15:11Summary and row level formulas cannot be directly composed.

15:14Summary and row level wait summary.

15:19Okay, and row level, so it doesn't like that I've not wrapped this in a sum, even though it's an aggregate calculation anyway.

15:31Uh lovely require use ag.

15:34Okay, now I have to set user ag on that

15:38Man, like I thought I instinctively knew Taboe calculations, but I can't believe I've had to like fight my way through this and then get to a percentage of a percentage of total calc.

15:47Just to do a percentage of total.

15:49Come on

15:51Oh, caused by semantic measurement label is required and cannot be empty.

16:00Alrighty, okay.

16:01Patience, patience.

16:03Wait.

16:04Inva input value creation of semantic entity of total calculator caused by semantic cap measurement label is required

16:11Semantic calc measurement label.

16:17Name?

16:17Does it mean name?

16:23I think that's what it is.

16:25I think that's actually what it meant and it just didn't say that.

16:28That was a really badly written error.

16:29Basically

16:30I was trying to create an API name that didn't in any way correlate to the name that I'd given it.

16:35So I had to give it a name for the API name to go through.

16:39So now I've validated that that works.

16:41That's now okay.

16:43Okay, that's fine.

16:45So now if I put the percentage of total on label we got percentage of total

16:55How long did that take me?

16:57That's so embarrassing, honestly.

16:58But I don't think it's my fault.

17:00Like um I've been using Tableau for a long time and that I really had to problem solve my way through that donut chart.

17:06Anyway.

17:07You'll notice a lot of sarcasm in in my tone.

17:13That is a way of giving feedback.

17:15It's a somewhat not constructive way of giving feedback

17:19But at the same time, I think I've made several points around intuition, and this is exactly what I mean.

17:25I don't think everyone who's going to be coming to Tableau next will be a Tableau user, but

17:30If you claim to be bringing the best of Tableau to your platform, anyone who's touched Tableau will know that it's a lot easier than this to build a pie chart.

17:41A lot easier.

17:42And so

17:43Again, this is still beta, but d come on, like this is a basic chart type.

17:50Percentage of total is gonna be something everyone wants to do.

17:53They shouldn't have to do an L O D.

17:55Like they won't know what an L O D is.

17:56They just won't.

17:58Like uh I know an LED because I've taught thousands of people what it is.

18:01But not everyone will know that off the top of the head.

18:04Anyway, so let's hit save.

18:05We've got a buy chart.

18:06Okay, that's one that one down.

18:09Many more to go.

18:10Maybe we'll just do like the first three.

18:12A bar chart.

18:13I think this is easier.

18:14This is much easier.

18:15So what we can do here is go

18:17to the top right select new and it just opens a new tab so for this one we're going to choose subcategories is nice and easy and then we're going to put sales we go put that on columns we'll do a nice sideways

18:27Bar chart, um there we go, beautiful.

18:30Uh can we sort this from descending to greatest?

18:33Yeah, we used to have them on the axis, but it's not available, but that's fine.

18:38Okay, can we stack this by maybe segment put the segment on color?

18:42Great, perfect.

18:43If I click on color, you get different ability to change the colors

18:48Um oh this is modal you can move it to where you want that's kind of nice um I did not did not even clock that so when you click on this it pins it there and then you grab this and then you can move it around so that's good

18:58Reset just resets to what it is.

18:59How does it look like to change color?

19:01Oh, they've got some good ones.

19:03Um oh they've got like some categorical ones as well and then custom

19:08um you can choose your your default uh style.

19:11So this is this is quite nice, nice and basic.

19:14Size, we can change the size of the bars.

19:16Yeah, we can make that all good.

19:18We can do absolute and relative, so that's kind of good.

19:21You can make them

19:22over stack each other that does not look good so we'll just leave that at that um again same with this you can actually move it around that's just a nice touch and labels

19:31We have no labels, so let's put the label.

19:34So let's get segment on label and then we should see that on the chart.

19:38Good.

19:39There is still no label option, so it doesn't look like we can format the labels, but that's okay.

19:43Tooltips

19:44Um obviously you get these tooltips.

19:46Um if I bring a segment onto tooltip, now am I able to edit it?

19:52Nope, I'm not able to do that.

19:54Detail, no, um detail is just a visual element.

19:57Range is interesting, range.

19:59What is range for?

20:00I've no what is range for?

20:03I don't know.

20:04Let's bring in sales and put it online and then say let's yeah, let's calculate the average sales.

20:10So hit apply and

20:12It is it gonna do it?

20:14Uh any anywhere today?

20:16Uh show the line.

20:19Measure.

20:21Oh it's the I think I chose the wrong one.

20:23It should I want to try and do the average

20:26Um what am I missing?

20:32Okay, let's remove this.

20:34Let's just try this again.

20:36Sales online label um value.

20:41average add line to average sales sales value sum of sales

20:50The average sum of sales.

20:52That yeah, okay.

20:53That add line to average sales.

20:58Is this

21:01Um hold on a second.

21:03Why is this average?

21:07Why is that average?

21:08That should be some.

21:09That should be some.

21:11That should never be average.

21:12No wonder I can get anything.

21:15But then even so this I think this this page is just frozen.

21:19Let's just refresh the page, see what happens.

21:21Let's just reload the browser.

21:22Does it

21:22Does it fail gracefully?

21:24Did it remember anything I did?

21:25No, I just reset it.

21:26Fine.

21:27Let's try again.

21:28I think something very weird happened there.

21:29So I don't know if I just lost the plot completely or if it's just a product.

21:33So let's go ahead and put the sales on there.

21:35To be fair, it does pick

21:37Sum of sales okay and what I tried to do is put sales online uh and there we go.

21:43I don't know what happened before but I clearly

21:47Mess something up.

21:48I don't know if it's because I put segment on colour that that threw that line off.

21:52So if I now try and put that on the line, it still works.

21:56Um

21:58But notice it's a lot smaller and I think it's doing it broken down by the um

22:08That is very interesting.

22:10So if I remove segment from this visualization and then I put sales on the line, we get a much higher average because it's looking at the whole bar.

22:17But if I put s if I if I hit apply and then I move segment onto color

22:22I would expect that line to stay the same, but what it's actually calculating is the average of each of the components, which isn't what I would expect.

22:29That line is not accurate.

22:31So then I say, okay, can I change this please?

22:38and add the line to sum of cells, the value is sum of cells.

22:43I want you to compute the average, but it's computing the average

22:48at the wrong level of detail.

22:49Okay, so that to me is a very clear bug.

22:52And that shouldn't be how it works.

22:54So yeah.

22:56Um yeah, like I guess I guess that's good.

22:59Okay, cool.

23:00So that's a bar chart.

23:02We'll call it so.

23:07Hit save

23:10This is wrong, objectively.

23:12Um that's not the number it should be.

23:14Just for argument's sake, before someone screen caps this and says Tabitim can't build charts, I will just leave it at the right looking number and we'll say no no trend lines if you have

23:24um segmented bar charts no trend lines okay um right two charts line chart this should be straightforward right like let's come on let's

23:33Come on, it's got to be the line shot has got to be the one with no bugs whatsoever.

23:37Come on.

23:38Um order date rows and let's do sales

23:43oh there let's do by weekday no day value by weekday and then um now what i want to do is rotate it so swap rows and columns there we go that's worked

23:58And let's try and change this by week.

24:00Let's try that.

24:01Does that look better?

24:02Okay, and then by month

24:06Okay, great, so we get that.

24:09Now let's let's break this down by segments, see three three different lines per segment.

24:14Good

24:16Alright, let's bring this down to quarter.

24:18The reason I'm just trying lots of different things is just seeing how it handles like that indecision.

24:24I'm not gonna mess around with trend lines again because I feel like that's burnt me.

24:28That's not really the purpose of this video.

24:30But that's fine, that's a line chart, right?

24:32Okay.

24:34So we'll call this line chart.

24:37So far, so good.

24:40Um scatter plot

24:43This is the one where I'm starting to feel like okay, I'm not sure this can be done.

24:47Um Okay, s so

24:55Oh no, scatter plot can be done.

24:57Absolutely.

24:57This is straightforward, right?

24:58So you get the two measures, sales.

25:03And then you get uh quantity and what you do is you put uh let's put subcategory on on color

25:12And then let's put custom on detail, that's about right.

25:19And then voila, we've got a scatter plot now

25:23Because the quantities fall into buckets, you can't have half an item.

25:26It looks like it's falling into sort of uh a slightly different bucket.

25:28So what we'll do is we'll go ahead and put profit instead of quantity

25:33And we'll get more of a scatter plot and then this is a classic thing you maybe have done in Tableau where you're trying to look for uh

25:42Where is it?

25:42Oh my god.

25:43Are tables not unprofit unprofitable here?

25:46Surely, surely.

25:49Tables are supposed to be the most unprofitable things.

25:52Um

25:54Oh wait, uh furniture.

25:59Furnishings.

26:00Are tables not in this dataset?

26:02This and of course it is.

26:03I've I've brought this from um from Tableau.

26:07Oh did they change it?

26:08Or am I missing something?

26:10Subcategory product name product name.

26:13Let's put product name on detail.

26:14Let's see

26:18Hmm, binders apparently.

26:20Binders that are really, really not doing too well.

26:24Um

26:27I'm honestly surprised by that.

26:29Um it may be because I have them so aggregated.

26:33So there we go, bookcases, tables, there we go, tables are the most unprofitable thing

26:38Because I had it down by customer, so there might be a customer who's buying a product that is more unprofitable.

26:43But there we go.

26:44Tables are generally the most unprofitable category.

26:46So that's all good.

26:47Scatter plot.

26:51Alright.

26:52Okay.

26:56Histogram.

26:57Oh this is more tricky.

26:58This is basically a bar chart.

27:00But what we're gonna use for this is we'll use the quantity um item 'cause that will make a good histogram, so

27:08Let's put this on columns.

27:10Now what you used to be able to do in Tableau is create bins.

27:14Um obviously that's not here yet, it's quite an advanced feature.

27:18So once you've created bins

27:20you would then uh use that to build a histogram.

27:23Because we don't have bins, um we're not going to be able to uh work with this but what we can do is we can turn the quantity into a discrete um

27:33uh element and then we can say um oh let's make this a dimension and then you go there you go these are the quantities

27:421, 2, 2, 4, 5, 6, 7, 8, 9, 10.

27:44And then you go put the sales, the amount of sales in each quantity of items that a customer buys, and there you go.

27:52So basically this is a row level look at the quantity column.

27:56And then what you could do to more easily see this is like say put an order ID on on colour.

28:01Let's see what happens if you put yeah, there you go.

28:05You get uh a sort of strange number there.

28:08So I think

28:08I think the reason uh why why is that okay we're only viewing the first 5,000 results it's obviously got a like a limit so we're going 1, 2, 3, 4 and 5 and 6 have been cut off

28:18So if we carry if we go back, then you'll see that actually, yeah, okay, it kind of works.

28:23So the first 5,000 rows ends up being somewhere through the fourth column here.

28:28So that makes sense.

28:29But yeah, there's our histogram.

28:31Nothing um nothing sensational about that.

28:37Um

28:39Bullet graph.

28:42This one is more tricky I think because what we're oh wait I keep saying that and then it's actually not.

28:49So with a bullet graph what we really want is a bar chart.

28:54Subcategory where subcategory.

28:57Sales.

29:00And then this is where we get a little bit, I think, stuck because

29:08I can only do lines.

29:10I can't choose the level at which they're working.

29:14So that that doesn't work.

29:17Um I like I don't know if I can put segment on fork.

29:22No, I can't put a dimension on this to sort of drive it.

29:24So I think this is one where we can't actually do anything.

29:29If you know how to bring those sort of analytics features, let me know.

29:34It's even not of like a hinted chart type here on the right hand side.

29:38So I'm just gonna assume that it's not it's not possible at this shit point as uh Salesforce would say.

29:43Alright, so we'll just leave this as is and I'll call this um bullet not possible.

29:53Alrighty.

29:58Packed bubble chart.

30:00We're very close, we're very close.

30:02We're going for 30 minutes.

30:04Um what is certain I would have done this faster in in Tableau itself, but

30:08Again that's not the comparison here.

30:10What's this like for a Salesforce user?

30:11And I think this is quite powerful um in many ways.

30:14So uh

30:16I forgot the chart, bubble chart.

30:17So bubble chart is basically a circle chart and then you put um sales on the size and you go and put um segment on the colour and then you get bubbles

30:30Now this is interesting.

30:35Let's say fit the entire view

30:38Okay, why why is it in the top right hand corner?

30:42I have no idea.

30:43Let's call that a bug.

30:45Let's go to the entire view and

30:48It's still quite small.

30:49I don't know why it's not filling the view.

30:52That would be quite nice if it filled the view.

30:54It's not really an option you get.

30:56Um all of them are stacked on top of each other, not what people were going for, I think, with the bubble chart.

31:03Um the circle is sort of somewhat broken behavior.

31:06Same is true with the square, um nice little animation between them

31:10you know, morphing that into another chart.

31:13Um it it sort of kind of works but it doesn't work at all.

31:18So um long story short

31:22segment on colour sales on size does not give us what we want and it kind of overlays them so that's not what we're going for

31:34Dot matrix, scatter plot, fine.

31:39Box and whisker, no chance.

31:42Tree maps.

31:44I wondered if uh let's call this um bubble.

31:52I wonder if

31:59We can use a bar chart for that.

32:01So no, you can't use a bar chart for that.

32:05I think this is gonna have the same problem as before.

32:06So if I bring sales onto size

32:11At least that does start in the right place and then colour Monta region

32:17uh uh the biggest one just crowds the rest.

32:21Which is strange and I can't change the order of these.

32:24So yeah.

32:25Um it is what it is.

32:26These are just uh not possible, at least.

32:29There is one here that it says um

32:32is possible dot matrix is the only one I've not done and obviously I've not built a table so if I just go to dot matrix what it wants is a dimension so let's put region on dimension

32:43And I guess because I've put region and region it makes sense, but if I then say actually put segment and region, then you'll get a dot matrix of cells

32:53I don't see the value of this chart, truthfully.

32:56This is just a heat map with like a nicer chart type.

33:01And so there were very some stylistic UIUX choices here

33:06But I personally would prefer this as squares.

33:09I would personally uh crank oh if you crank it all the way up

33:15It doesn't quite work, so uh they overcrowd each other.

33:20So you have to basically go for 99% of the cell

33:26And then it starts to kind of work a little bit better, but it's not quite perfect.

33:30So yeah.

33:32Listen, I think we've done enough.

33:34Uh

33:35charting um the one thing I have not done I can't believe I've not done this is a table let's say let's build a table come on let's let's do this uh we can just be in customer ID

33:50Data source category city.

33:55Okay, that's it.

33:57Nice clean table.

33:58Nicely done.

34:00Okay, so we'll call this one dot matrix.

34:06Hit save.

34:09And we'll call this one a table

34:20Now that we've done that, go back to our workspace.

34:23We should have a whole host of charts that we've built.

34:27And yeah, you can see them.

34:28Nine visualizations, nice and nice and clean.

34:31Um and there we go.

34:33So my takeaways, we've been at this for 34 minutes, makes sense to stop here for this video.

34:38Um

34:39These were basic charts.

34:41Obviously it's clear that they're still adding charts to this platform.

34:45The one problem I think I have with this is um

34:53I think for some of them, for example the pie chart.

34:55I just want to be able to build the basic pie chart before I go building a donut chart.

34:59Like the donut should be an option in the pie chart, but you should have built the the pie chart first, then enable the donut feature.

35:06So like

35:06order of events and priorities to me they're just completely broken.

35:10Um the bar chart

35:15being incorrect in that instance where I had a reference line, uh where it decided to do the reference line based on the level of detail of the Viz rather than the actual charts that it's supposed to be presenting.

35:25with no way of changing that.

35:27Normally in Tableau you can just say, look, do it at this level if if that's wrong and it it does it at that level.

35:33with no ability to control that I kind of feel like that feature is not complete.

35:37You can't say that feature is done if it doesn't do the right thing when you create a scenario where it should still work.

35:43So like that should be the standard.

35:45You should be testing every single possible

35:48um sort of permutation because if the chart is wrong, it's wrong.

35:53No one will ever trust it.

35:54So you can't ship things that allow the user to build things that are wrong.

35:58It's just not not a good place to be

36:01So yeah, I think I've spoken enough about the bugs, I've spoken enough about the experience as I went through this.

36:06So you hopefully got an idea of that

36:08Um now before we we come off very briefly I'm gonna set up what I'll do in the next video which will probably be in like today is actually Thursday slash sorry today is Friday

36:18Um tomorrow in this video sense will be um let's say Sunday So on Sunday

36:27Tomorrow, according to this video, um, I want to come back into the semantic model here and build a metric.

36:33So up here you have the ability to add a metric.

36:35Obviously, you can add parameters and other things

36:38If we go in here, you'll see that this creates what is almost identical to the Tableau Pulse interface.

36:44In fact, I would say metrics in Tableau Next are identical to Tableau Pulse.

36:49metrics which is so confusing.

36:51I just think they should call them all Tableau Pulse rather than having two things do the same thing built in different parts of the ecosystem.

36:58It almost felt like Tableau Pulse was a beta exercise for this and they're kind of using that to learn

37:03a bit about the experience.

37:05So what we'll do is we'll do this in the next video.

37:07We won't do this today.

37:08We'll go to metrics, then we'll go and build a dashboard because we want to be able to use these metrics inside of the dashboards.

37:14Thanks for watching.

37:14I'll see you tomorrow