Go! Tableau

At least, hi everybody. So yeah. Welcome to Go Tableau. As Vicky mentioned, we've got a lot of content to cover today. There's a lot of features in Tableau that we can dig into. And we are gonna try the best to cover as much of it as possible the idea of today's session is really a kind of zero to competence or being able to get started. I've been using Pamela for, for LBS or something now. I'm still learning new features all the time on Tableau. So we won't cover everything, but to be honest. But we will hopefully give you enough, information and all that, enough, expertise to be able to get up and running with the tablet desktop products connected to data, start creating charts, maybe add some calculations yourself, and then build a dashboard before publishing it to your, type of cloud or type of server account to share with your colleagues. So that's the goal for today's session. I'm getting up on ninety minutes, so I will get into it again, I'm Max. We've got Zoltan here as well. Zoltan loves coffee, so he will be absolutely on it in the chat. So if you do have any questions, he will be or attentive and get to you ASAP. So, brief introduction about interworks I hope that you know us already, and you've heard our name at least. We are a global Tableau partner, and we're the first school partner of Tableau as well. And, and, we have about two hundred fifty consultants, around the, the, the full corners of the earth. We also have, world famous blog, special items and everything data analytics focusing quite a lot on Tableau, but there's other things in there for our other partner products that we align with too. So go have a look on there and some really good content, if you're having a coffee break in the morning and you want to flip through some daytime analytics content, then go, have a peruse. So what we're gonna be covering today, is the tablet platform to begin with. So I've got a few slides here. We're gonna kinda of our, our PowerPoint deck sooner rather than later. But we do have, sort of contextualization of the tablet product to begin with. So that people understand exactly what Tableau does and demystify a little bit of the terms around Tableau. What do we mean when we say Tableau? Do we mean the Tableau desktop product? Do we mean the server product? So there's a little sort of preface there on, on the difference of technology that we have. And we're also gonna introduce this concept of self-service analytics, thinking about how things might have been done probably years ago now still exist today in certain organizations that were maybe just catering for ad hoc reports constantly and we don't have this concept of self-service where we can actually find the answers ourselves. How does that compare with the kind of more modern data stack? We're then gonna dig into Tableau as a tool. So thinking about the Tableau essentials, how can we get the answers to your questions in Tableau from the perspective of the business user, someone who's really not, a data professional. Right? So we're gonna be rolling through these steps and we'll go through these in a bit more detail. Vicky has added into the chat. The link to the starter workbook, if you do want to play it along here, This is a Tableau file. So if you have Tableau installed and you double click this file, it will open up in the Tableau desktop product. And, we also have data file in there that we're going to be connecting to as part of our exercises. So this is the global Superstore. We'll see it shortly, but it's this is like a, transactional online sales data source based in Excel. So If we have those, then we can download that file from our, folder that should be shared with you in the chat, and you can download these or the, the the zip folder that's above it. And then once you have that downloaded, you should have essentially a zip folder like the in your downloads folder, if you're using Windows and the equivalent on Mac, and you should be able to unzip that and you're gonna be good to go to play along today. As I mentioned at the start of the call, there is a lot of content today. So the speed might be a little bit quicker. So the, so the the recording be very useful. Also, when we get into the Tableau, tool, there's gonna be captions on various different slides. So if you want to come back to it later, you can do. So don't worry too much if you do a little bit behind. It's kind of the, the head of concepts that we're gonna be interested in, in kind of communicating today. So let's discuss a little bit around the landscape of Tableau. Where did it come from, and and what's the kind of philosophy of the products? Pat Hanrahan in two thousand and five, he was an Oscar nominee or an Oscar winner, I think, as well as Chris Stotta, within the, university of Stanford, I think, I think you're with Stanford. They created the database query language that underpins Tableau is the SQL growing language. So when you drag something in Tableau and it brings you back data, that's essentially writing SQL for you in a visual fashion, which is where the term VISQL comes from. This speeds up our ability to analyze things quickly. We don't have to type out syntax. We don't have to get it wrong if you're like me and then spend some time debugging it. This is a nice quick, easy way of aggregating data and pulling data out in the format that you want, in a very intuitive, hopefully fashion. So this is the same lap that created, the Viscural engine is also used for, Pixar. So, it's also kind of the the building blocks of, our friends at Toy Story and others as well as, this is the same lab that attached Yahoo and Google as well. So I don't know. It's quite cool, but if trivia about Tableauk before we get into the the nuts and bolts. Second thing I kinda want to produced here is the idea of democratization of data. So as I mentioned, Tableau should be very intuitive. It's a nice drag and drop interface And as soon as it's not getting familiar with your own data sources, you can create charts that have impact very quickly. Previously in the old world, kind of waterfall development, this would be a function of IT. So I had a question. I would have to go to somebody like Zoltan if I'm a business user in Zoltan's in IT, and I would have to say, What are my, what was my revenue in the city of Edinburgh where I'm from over the last four years by month? That's almost a better question than you're likely to get if you're in IT at that point. So it's got, it's got quite a few criteria that you can understand. You may get something like what's the revenue here over time? In which case you gotta go back and you've got to ask, ask more questions of what they want to look at. And the idea of hello, and the idea of democratizing data is the result and, or sorry, me, in this case, can answer those questions myself just by being provisioned to the data that I'm allowed to see in a well governed format. So this speeds up the, the, the life cycle of, analytics development considerably. We can just answer our own questions and we can get those and then we can follow them up. And then we can do all of that ourselves rather than having the communication back and forth. So this is kind of baked into a lot of analytics products and almost every single analytics product from a visual front end perspective in the modern world, but this is something that was really pioneered by Tableau. So this is their, their, their opening mission statements, back in two thousand and five. So this is kind of baked into everything we do in Tableau. So we'll see today that there's very little requirement to have coding understanding. There's very little requirement be incredibly technical when using the tool, which is why it's kind of, the the, the focus is really for the business user. That's it. It doesn't bottom out when you get, when you you can't just complete Tableau in a couple of years if you're technically minded. There is a massive learning curve that keeps going. So the initial learning curve is quite short. Hopefully, we'll get through a lot of that in nineteen minutes today, but keep going with Tableau and you'll be able to create things that, can rewire your, your, your colleagues. So Tableau is a data visualization platform. So, again, when we say Tableau, there's a few different things that we mean. I personally usually mean Tableau desktop, but again, I've been using it for quite a while. In your organization, when you say Tableau, you probably mean the endpoint that's at the end of your URL which is likely to be Tableau cloud. So, I am looking for this information. Have you tried looking on Tableau? That is probably discussing Tableau cloud in this case, which is the sharing platform, for Tableau, which is based on the in the browser. So There's many different elements of Tableau, and the first element is our ability to prepare the data. Our second is to be able to analyze that data, to build those charts, and those graphs, and answer those graphs the third piece is to be able to share that with our colleagues. And then finally, we want the ability to govern that data. So we want some people to be able see some data sources, other people to see other data sources. We want that to be intuitive, and we want to have a flexible commissioning model so that some people could do things that other people can potentially and that works for most people again in Tableau's eyes. They want this platform to work in as many different scenarios as possible. Which allows, or or leads to a very flexible and quite granular permission structure that's evolved over time. We're not going to get into permissions at a huge amount today. We're gonna be looking at the first three elements of this, this, this slide. And we're gonna focus on the, the parts that I've folded out here because I really think that these are the biggest parts the others are used in other scenarios. There are scenarios where the other on mold and pieces are, are very useful. So tying off with our, preparation of data. So we know where the data is. It might be sitting in Excel. In our example, But it's not necessarily perfect. There's a few different things that we can do to change that, and we're gonna look at them in our first example today. The first part is preparing data is our ability to actually ingest that data into Tableau itself, which is where the connectors come in. Those connectors basically allow Tableau's VISQL language that we mentioned earlier translate into the language that the data is stored in. Now for SQL, if you're using SQL database, for instance, that's one type of translation. So I want to sum up my sales by Citi. That is then going to be converted into an SQL script that's going to run on the data source. If we're looking at, excel, which is very different. It doesn't run on SQL, then there's a different language that Tampa has to, apply those. So we're gonna get into that a little bit more detail later There's also some more things that you might have to enter like your credentials to a database for instance that you don't have to do on Excel. So there are some differences there, and we'll see the different connector options in Tableau. We also have Tableau prep builder here, which we're not gonna get into more detail after this slide, but Tableau prep is a, a last mile data prep tool. That allows us to perform complex manipulation of our data sources before we ingest it into Tableau. Now that runs as a kind of workflow so you can drag and drop a different task into that workflow engine, and you can click on run at the top there, and that's then gonna sequentially run through those tasks that might be filtering the data, joining it to another data source maybe pivoting it in more complex sort of manipulation transformation, jobs. And then, finally, we spit that out of the data source that's then ready for example to ingest So capital prep builder, if you're interested in more of last mile data prep concepts for using lots of files, often it can be a really useful tool for, for ingesting complex data. But generally most of what we do from a Tableau connector or or ingestion from Tableau point of view is perfectly available within the core Tableau desktop. Product. So a bit of a special case type of crypto dataframe. When we move into the analyze space, if we're looking to analyze our, our or or data once we have ingested it. My favorite here is Tableau desktop. Lots of other people might just simply use the web editing functionality within Tableau Cloud. So which kind of mimics the desktop product, but we're going to be looking at the desktop product today because that's got full functionality, of, of all the formats in the small sort of, more fringe functionalities available on Tableau desktop and about ninety six percent of that, I would say cautiously is available in the web edit. So this is gonna be available on my local machine. I'm gonna be running this from my laptop, but there are there is the ability to run a lot of this directly within the browser itself, within the editing functionality, of Apple Oclane and Alba Sogo. Then when we get to sharing ODEtha, Sharing is once we finish a dashboard or maybe we're halfway through and we want somebody else to peer review it, then we don't send them an email with the file attached to it. We usually publish that to the browser. We enjoy the governance and the commissioning and the secure way of interacting with that data within the browser. And we might send somebody a link or say if you tried looking in this folder, I've made a new dashboard for you. So this is what we're gonna be focusing on in the last five minutes, we're going to be publishing up to a Tableau server, instance or a Tableau cloud. Tapau clouding Tableau server are almost interchangeable from most of the the conversations, we'll be having. So if I say one, I mean, of, Tableau cloud is a software of the service version where if Tableau server Samsung Campbell Server is one where you calls the infrastructure yourself. From a front end point of view, they're pretty much equivalent here. And I've also noted some other, some other tools that are available within the Tableau ecosystem here as well. In terms of governance, again, there's a Tableau catalog, which is, which is available, from, from Tableau But generally, governance means lots of different things to lots of different people. The permissioning model of capital cloud and capital server allows you to surface the right data to the right people, which is probably the main the main job of governance in the data analytics world. So what we're going to be looking at within the temporal essentials of all, we're going to be connecting to data. We're going to be exploring those data connections, as I mentioned previously, and looking at the different options that we have available to us. We also have data prep or data preparation. So if we need to manipulate the data before we perform our analysis, because it might not be in the exactly right format for the analysis that we're trying achieve, this is where data preparation comes in. Once the data is prepared, it's all perfect, it's all squeaky clean and shiny, we then get into creating our charts. Bar charts, line charts, pie charts, don't have charts. That's all great. And absolutely Taddles button button. So we're gonna be getting into that once we've prepared our data, we're gonna build probably five or six different charts at least, and that will then give us the ability to enrich those charts with calculated fields. I mentioned at the start, this is not something that we need to be particularly worried about from a sort of technical point of view. Table syntax per calculated fields doesn't really go past these Excel type. Syntax. It's quite similar and kind of akin to Excel. I'd actually suggest it's a little bit easier than than Excel, generally. I don't know if it's all time. Do you do you agree with that? Excel versus Tableau, since the same time sequence? Actually, I would say one hundred percent because mostly, simply from the fact that, it's not in one line. It it makes, you know Right. One hundred percent agreed. Yeah. So we've got lots of flexibility and tablet calculations, excel scares me a little bit somehow, but, yeah, I don't know if there's I should be admitting to that. The, interactive dashboard functionality is where we bring all of this together. So we've made a bunch of different chunks and we want, one unit, one interface so that our users can, explore the different chance together. We can click on one thing and it's gonna impact the other chance in a way that we choose. So this is where the the rubber really hits the board. This is where we can create analytics tools for people to go and answer their own questions. And then finally, once we build our dashboard, we can then share this with our team, by publishing it into the browser. So first of all, data connections. So, let's go back to the beginning, and that is our connecting to data. Before we analyze our data, we need to get it into Tableau. That can sometimes be the biggest sticking point, for people that are first getting, getting to grips with the tool. So Tableau has a time of creating this live ninety plus native, data connectors. And we by native here, we basically mean that there's an option in the list that says, I want to connect to Salesforce or I want to connect to my sequel or, or even postgres sequels, something like that. The lots in this list is mainly plastic and options in this list, and that is growing all the time. Every time there's a new work in the catalog, there's, there's the more added to go to the list as the, the, the, the, the, the data stacks of Pappl's customers changes and grows. So it's an ever changing list and these allow super simple connectivity to your data sources. So we're gonna start with a quick exercise here. So as I mentioned before, we should have the devon loading data if we're, anticipating playing along. We should have the data already is on our local machine as an Excel file for instance, and we can just pop into the, folder structure that we downloaded it to. We can double click on our version twenty twenty one dot one starter. Go Pablo. I've already opened that up here. So you should see either something like this. Let me just make this a little bit bigger without maximizing it so that everybody can see. And this is the Tableau desktop interface. Now you might see some thing like this to begin with without all of these, squares in the middle of here. This is the tamil start menu. This is equivalent for me of the Excel start menu. If you don't click on a Excel file, if you just open Excel from your start menu on Windows, then you'll see something that is equivalent in functionality based. It'll ask you, do you want to hop into something that you did recently? Which is basically this main section here. We've got lots of different, workbook year that I could hop back into that I've been working on. I've also got the ability to connect to new data, which is what we're gonna be doing here. I can add a data source so that I can start analyzing my data, or I can look at the right hand side, which is kind of more intimate driven here, which updates quite regularly say maybe you want to learn more about relationships and tackle and building your own sort of data, models and that kind of thing, or of the new features that Tableau is coming over that you might be interested in over a year. They might also drive floppy tickets to the Tableau conference of any Russian and software down at the bottom as well. So, again, depending on what you've clicked to open it, you may see this here, or you may see this. If you do see, this one, then just click on that Tamble button up here, and you should see the different tabs across the bottom very much akin to Excel here. So these different tasks are indicating the different exercises we're going to be rolling through in today's session. And if you don't see them, make sure that you have double clicked on this starter workbook TWBX, and you've opened that file rather than just opening up to have a little one. So So if you do see this, connecting to data is our first tab here. Each of these tabs exists as a separate atomic chart that we haven't built yet. We're gonna be getting into these in, in sequence throughout today's session. As I mentioned, we've got the captions in the top right here. So you can always come back to this later to follow those instructions. And when we finish, we can see if this is another one, another another file, and we can always sort of have that for one to come back to this and run through the exercises again. So in the first section here, we've got our connect to data questions. So we've got our, connected data in the top left here. We've got a bunch of directions here where we're gonna be connecting to that excel file that we've got in our local repository that we've downloaded. And I want to ask, Tableau to ingest the data from that Excel file into my tablet topic of it. So I'm going to click on connect to data here. This is the exact same thing as if I go into this sort of data source symbol that appears, obvious add a little cylinder with a plus, and I can then select from this list, Excel. So there's file structures our file objects that I can go and connect to. There's also server objects, and as you can see more is highlighted here where we see those ninety plus different options that we have to connect them to data sources, on, whether on the cloud or on a on a server that is accessible to us. So I'm gonna click on Microsoft Excel. Again, if you click at this point, and this is the same as if you click on connected data here, there's a few other ways of doing it data and new data source. So Tableau allows the method that you want to kind of use. There's lots of different ways of kind of performing the same, the same action in Tableau. So I'm gonna click Excel file. And I'm gonna go to my gold tableau folder, and I can see in here because I clicked Excel file, My Windows Explorer has still at this time to show me all the things that are of type, XL or XLX, and so on. So I'm gonna double click on that. That's then gonna open up a new window of Tableau. So I've moved away from the analytics window, the chart creation side of tableau, and I've moved into the data source manipulation side of Tableau. So Tableau's picked up that Excel file. Over here, I can see in the connections in the top left. This is pointing to an Excel file of this name and this folder of my, of my local machine. And I can change that by editing the connection or maybe renaming it if I need now we're creating a data source in Tableau that I want to use for my analytics. And because there's only one tab in this Excel file, I can see here that I've only got the orders file. So Pampers actually drag that into my canvas here by default. Now if you've got an Excel file with multiple different tabs, you might see something like this to begin with with lots of different tabs going down the left hand side likewise with databases that have multiple tables in you'll see a different interface here. So what we need to do here is we need to pick up the object of the table of data that we're interested in analyzing, and we can drag it anywhere in this main canvas. And we'll get back to that point. Now this is gonna basically create a library of data that we're gonna be analyzing in our charts, and we can have different libraries of data, or for all the different sheets or different dashboards that we've got within this workbook. Think of it as an Excel file. It's got lots of different tabs, we might want to connect to different places in order to perform in analytics ourselves. So we can give this library of data a particular name. I'm just gonna call it orders because that's nice and easy. That's what I'm gonna be able to see in my data menu when I get into the Tableau analytics and visualization interface. Down here, in the bottom half of my screen, if I make this a little bit bigger, we can see that we've got all of the columns listed out on the left hand side here, which makes it nice and easy for me for instance, change a name here. So I want something to maybe show differently in Tableau from my developers in Tableau than I did coming from that raw Excel file might want to change things about in terms of the metadata here, which is the data about the data. That includes names, and aliases of the different values that we might have, we might want to do a finding place in here, for instance, also it includes the data type. So maybe I've got a year over here that is actually a number, which is indicated by the little green hash symbol, Maybe I want to make that a date so I can try and push that into being a date and see if Table manages it, and assumes that the year is actually the year. Yeah. So that's right. That's that's actually got that spot on a little bit dubious that it would. So here we've got the year of the order date doesn't actually seem spot on. But we've got the year of the ordovator. We can always hit control zed to go back to, to, to, undo the changes that we we've done. So this is the first start of our data preparation piece. We've connected to the data, and then we might want to prepare the data by manipulating some of the metadata in this file that it's nice and palatable for our end users to be able to start doing that self-service analytics if we want to hand this over. So In this section, we've, also got, again, the names we can scroll across. Tamela's just going to be querying the top one hundred rows here. So we're not seeing all of the data we do want to see all the data in that Excel file, I can go and look at my connection on the left hand side and I can view the data in this window, and this isn't gonna have any of the changes that I might by, this is kind of the raw data from Excel, in this case, that I can just go and check that the levels of granularity are correct. There's no kind of hidden roles columns that are causing any issues in here, and then I should be, off to the races and and good to go. I can choose to extract this data for performance potentially, or I can keep this as a live connection to the Excel files. So if I update the Excel file, it should just feed through straight away. When I click refresh, I can also add a filter on the top right here to say I'm only interested where the country equals UK, for instance. So myself, I might have all the data My TableO might have a subset of that data. We're gonna see filters a few times throughout today's session. So this is kind of the first place we can apply filters so that nothing else is gonna filter into that data source as we go. Any SQL people out here kind of like a where clause. So now that we've got our data source all prepared, everything was good to me. I'm gonna hop back into my sheet, which is my analysis, my sheet one, my first exercise, I'm gonna check that everything is looking as it should. So here, I don't see any red exclamation marks, which usually means I've messed something up. I've done something wrong if you do. Then you probably usually right click and edit the thing that's got the exclamation mark next to it and it gets you into that window. And we can just look through the directions on the right hand side to make sure that we've done everything that's been asked dollars. We've left clicked on data from the toolbar. That's fine. We've created a new data source. We've connected to the Excel file. And we'd navigate it back to connect the data, which is the sheet we're up. Now finally, we want to check that everything is flowing through correctly. So we've got, the request to drag sales to columns, sales role by default be aggregated as a sum, and then we want to drag segment to rows. So just to that everything is looking good and to introduce ourselves to this data window on the left hand side. This data window again is kinda like our library of different columns from that cell file. So if I want to have a look at that raw data, I can always click on the little raw data symbol that you'll see cropping up every now and again. And we can see here if I make this a little bit bigger, we can see that, for instance, category aligns with category here. So our raw data is gonna have these values in it and our aggregated or our, metadata layer is gonna just that category that I can choose to use to build my visualization. As I scroll to the right, I'm gonna see more new numerical values like our sales and revenue and to call that, and I can see the shipping cost and the quantity of items that was purchased at any one time. So we've got all the data to be able aggregate and to, group our data together to draw charts. So as instructed at the bottom of exercise one, I want to drag sales onto columns. Now, everything that I drag into this canvas in the middle here, before I do that, let's just show you what I mean by the canvas. So this is specific to sheet one. So when I move on to sheet two, everything within that orange square is going to be, is going to be new. It's a new tab in Excel, essentially. I do have the connectivity capabilities over here, so I'm still gonna have this library of data to be able to play with, but I've moved on to a separate chart essentially. So my first chart, I'm gonna bring in sales, and I'm gonna drag that onto columns, as I said before, and what columns is gonna do is it's gonna distribute that data across the page. So by dragging on to columns, I'm gonna be distributing that data across the page. And if I drag it onto rows instead, gonna be distributing that data up and getting the page. So that's a little bit nebulous at the moment. I'm gonna show you exactly what I mean by that. So bringing in sales and dragging onto columns, we can see that we've got that distributed across the page, in this case, as a bar. It's quite a boring bar. It's just one bar, and that bar is giving us a value of about twelve and a half million dollars like this in this case. So we're summing up as we can see over here. When I drag that green object into the view, we're seeing some of sales sales is down here. That's what we've pulled in, but we can see capital has applied an aggregation. It's applied minimal sensible aggregation in this case, which is just summing up the sales. So it's totaling that column that we saw in our Excel file. Now I can change that if I want to maybe take the average value. I can make it the median value you or the maximum, the the biggest one or the smallest one. But generally, I'm gonna stick here with my sum of sales. And if I was to drag that instead onto the row shelf, I'd remove it and maybe pull sales onto the row shelf. Remember I said before I seen a distributor up and down the page, We can see that bar is flipped ninety degrees, and we can see that same twelve and a half million. That's gonna be going up and down the page instead of across it. So this is, again, green, and green in this case is going to be aggregated. This is Pablo's way of, of showing us that this is a measure. Now, If I am to drag instead this segment, which is as for on, question four here, I want to drag that onto columns I'm actually doing something that isn't gonna be aggregated in terms of the bar because this is blue. So there's a fundamental difference between what Tableau how Tango displays the two different types of data. So segment when I drop that in is gonna show the the distinct segment names that I have in my underlying data distributed across the page. Likewise, if I do that on rows, it's gonna distribute them up and down the page. So this is the building blocks of Pablo as a visualization, inquiry engine. Again, the difference here, if I go blue, I can see the blue objects up here. These are going to allow me to group my data. And if I just drag in sales as well, I should see a nice vertical horizontal bar chart. So I've got two of these different ingredients in my view at the moment. And the difference here is that one of them is blue, which is gonna allow me to buy a grouping object. So this groups up my data. And we can see here the tableau when I drag this onto rollers is going to distribute these out as labels. If you remember two things about blue, it's a grouping and it's gonna give us labels. If we conversely think about what the green object is doing, these are usually numbers and this is going to aggregate. And by aggregate, we mean totaling up or averaging or taking the biggest to taking the smallest my sort of maybe quite what I way of thinking here is that if you take lots and lots of different values, aggregation turns them into one value. So how does it do that? You've got a choice, what logic to use. You could add them all up, you could average them, you take the middle one, you could take the biggest one, you could take the smallest one, so by default, we're going to be aggregating our sales in this case, and and tablets used to sum that. We've proven that we can change that. We also, when we drag this onto our rows columns, we're gonna see its scale axis. This is because these are continuous, but they exist within these green objects exist within a continuum. So in this case, we can see that that is our, our scale access. So consumer corporate home office, they exist as distinct values, whereas sales exist somewhere in this, this this continuous, access essentially. So it's gonna sit somewhere on that axis and, and, and that is the difference here between a green and a blue object. If we don't remember anything about Tableau, other than one thing in this session, that's probably the most fundamental thing that is worth, really, getting the team stuck into different in green and blue. So I'm gonna pop back into my PowerPoint, unless there's any questions to solve time. Everything being answered? Nope. No questions so far. No questions. Cool. Well, got an easy day today. Yeah. You're doing really well. Feel free to, to best results out if you do have any questions on that. So the next step, data preparation, we've kinda covered a little bit of this already, but we're gonna dig into in more detail, in this section. So bad data equals bad analytics. Right? Some of someone once said to me, the only thing that, worse than having no reporting is having reporting was wrong. And often that can the the reason for that wrong reporting or the number is not being correct is either assumptions have been made or the data is incorrect. That is underpinning it. So understanding and being able to visualize the sort of profile of that data quickly, like what we just did there, is is fundamental in understanding our underlying data model rather than, you know, writing queries and, and, and see whether your putting it all in Excel and and scrolling up and down to try and find any errors. So if the data is close to being good, this is a very rough rubric. But hopefully you'll accumulate here. If the data is close, but not quite perfect, then, then tablet desktop is probably okay. In order to kind of make those changes. Maybe you just need to add a little filter here or apply a little alias there to kind of plan and replace some values, maybe filter out some null values in your way. If the data is not great at all, you might have to do more heavy transformation on top of that, which is where that tablet prep builder product I mentioned earlier can come in. Finally, if, if the data is very much not perfect even then, then you might have a bigger data problem, you need to go back to source to to fix that and and and and prioritize some of the understandings. So cleaning up not great data can be done very easily in tablet desktop, buying filters, maybe adding joins. There's not something we're gonna get today there is possible to do. And if you've got multiple different data sources that you want to combine in a more complex fashion, and that can that some of the combination and transformations could be clean up processes, then ample prep is a good way of, of doing that. So a couple of quick things that can sometimes be a little bit different in the excel world than the camera world. In excel, we like to read things from left to right quite often. So you'll see more columns in then classically you'll see in a database that's been built or a data source that's been built for a beyond tool like Tableau risk competitors. In Excel, having these as different columns means that the way that we're gonna interact with it, and Table is gonna be a bit different. Instead of the sales column that we saw before, we might have sales Q one twenty twenty, sales Q two twenty twenty, sales Q three twenty twenty. And that means we have to bring them into our onto a row shelf, onto a column shelf for every, quarter that our, our, data is updated for, which is less than ideal. A better way of doing this is to pivot the data. So this is a concept that's quite specific to, to, to data manipulation processes. And this is the same data or the same row of data over here is five hundred four six five three four five and two four six. And they're going down the page here instead. So we've actually taken this header set, and we've turned them into a column. And that's just flipping the data around a little bit. So it's a classic transfer that we might need to do before using a lot of Excel files. One sort of guideline that I can attend to tend to to evangelize sometimes here is that if in doubt, if you've got a very wide data source that's quite short, then sometimes kind of pivoting that round, to be, instead, a bit narrower, but longer. Is usually better for analytics. So if you have the same metric multiple times for different values, it can be something better to pivot that round. So, again, that time concept is something we see quite a lot. It could be you've got them split out by product and excel, so you might have your furniture sales and another column giving you your, technology sales. Because they're both sales, I would recommend pivoting in that example. If one of them sales and monuments profit, I wouldn't pivot because they're two completely different metrics. Hopefully that makes sense. So data preparation, again, sometimes we'll see something a bit more sort of flashy in our Excel in in our excel, example here. So we also go on to see these kind of subheadings or the common and excel document these are gonna cause Tableosome regime when it tries to ingest. So really the raw data is perfect for Tableau. If we've got something like this, then we're gonna have to do a little bit of work in order to prepare that data. We don't necessarily need to change the structure in Excel. We're just asking Campbell to adjust the way it interprets it. So Here, we can add calculations potentially in order to, get to our pet dose of biology, your physics here. And then we've pivoted around the month so that we've got our month as another column instead of separate columns, and then we pivoted out the number of students here we've got our five hundred. So again, lots of that we can do here in terms of data preparation, but generally, Tableau or Tableau Pride are are excellent ways of doing that if we don't have access to that upstream source. So we're gonna do a little bit of light preparation, not for thing on, on pivot things the way it's a bit of a complex example pivot teams took me a while to get my head loaded. So we're gonna be thinking about just basic stream manipulation in, in in this section here. How are we doing for questions all time? Have one question. I answered that. However, there is Fatima who's raising their hands. Is it possible for you, please, to to type your question in the chat? So, so that we can answer that. Thank you, Liz. Thanks. That'd be great. So while we, while we get that question through, and we move on to exercise two here. And exercise two is going to, allow me to manipulate what we have as product name, which is very, very granular. So if I drag my product name onto a raw shelf, remember that's gonna distribute it up and down. Yeah. Exactly. So up and down the page here, I can make that header a little bit bigger, and I can see exactly the unique values that I have within that product name column. And we can see here for the key insight that we have, for instance, a, we have a little comma in some of these. If I scroll down a little bit with the zoom, can see this echo hole punch in the middle of the screen there as, header level. It also has a comma for the clear variety, a comma for the durable variety. A comma for the economy and recycle. So I might want to just roll up all of these alcohol punches, the o g one, as well as the different values that we have after it. And I might want to roll them into a subcategory or maybe a category relative to its model. And I want to pull all of those together. So I don't want this low level granularity. I just want to strip out that value of aqua three hole punch. Before the comma. There are lots of ways we could do this. We could write about the syntax, to find that first comment and take everything before it like you could do in Excel, for instance, In this case, what I want to do is I want to leverage Tableau's kind of business friendly user interface in order to get, to that place. Without writing any code at all. So if I go to the product name that I dragged in, I can click on the little down arrow and this allows me to kind of manipulate the metadata that I've got, manipulate this column, do something to this column. Lots of things I could do in here like renaming it I could duplicate it to create a copy of that column and then start messing around with that copy. I use that quite a lot. And you can also do things like transform that column. So here, I might want to transform this column by splitting it. So I want to take echo three hole punch comma clear and I want to split that. Now, I don't want to split it based on the six character. I don't want to split it based on the, the first space or anything like that. I want to split it based on the first comma that I get to. So if I click on splits, tap those basically just gonna say, it's gonna pass all of those, and it's gonna say, yeah, okay, I'm gonna guess and say, let's use the first I think probably the first space or something. So let's have a look what Tamela does if I click on split. So I can now see I've got split one and I've got split two because it's highlighted both of those, I could just drag them into my, into my row shelf and I can see new columns in my data to see exactly what Tableau has done. Something a little bit weird. We can probably dig into what it's done in more detail if we wanted to, but it's not satisfied what I asked it to do. So or or why I didn't actually ask you to do much than specifically other than split. So I can be a little bit more specific. I think this first one has actually given me the first space But look at things, everything is there. Yeah. It's just not the first space. And then the second part is giving me a number after that space. So that's not And I'm gonna highlight here the most important thing in Campbell for me if you make as much mistakes as I do, which is the back button. So I'm just gonna click that back button or control z on my keyboard, that's gonna undo everything, and I can even undo that, that transformation that I applied that split. So now I'm back to where I was before I made this I'm gonna right click on my product name, go into create note, go to transform, and I'm gonna click on custom split instead. Now custom split's gonna allow me to be a bit more specific here instead of split. So transform custom split and I get a nice a window where I can be a little bit more, specific. So here, I'm gonna ask Tableau to split based on the first comma rather than the first space there. I'm gonna ask to split off the first one column. So I'm gonna ask for that first one. So the echo three hole punch, that's all I want. So clicking on okay, I can also ask for the last, but all of them, but in this nice user interface, I'm gonna split that first one off, and now I can see my product name split one not perfect name. We're gonna change that in a second. Now I can drag that just before product name maybe, and I can see exactly what that's given me. So in this case, that looks really good. So I've got Apple three hole punch, and I've got the various different models underneath that. But now if I was to drag in my sales, so I'm getting my horizontal bar charts here. If I remove my product name, then I'm taking the views at lower level of granularity, and I can see that my echo three hole punch. I've lost it, whereas a echo three hole punch here has all of those different products sitting underneath it. And I can even right click on that. I can view the underlying data, and I can see that I've got various different product names in that underlying and I can kind of debug in doing in that order to check, that that's happened in that aggregation. No. I can sort that descending. So I can see that my Cisco smartphone is the highest selling product at the, kind of, the, the product name rather than the model name. And I can see that I've got two thousand two hundred and eighty seven different rows. So that tells me that I've got two thousand two hundred and eighty seven different products. And if I was to drag in that product name and overwrite that product. I could see that I've actually got three thousand seven hundred and eighty eight. So we'd expect there to be more of the kind of more granular product names here than the high level products you're getting hitting controls out a few times. I could get back to the view, greenhouse degree in the exercise. Hopefully, that all makes sense. So a bit of lightweight data preparation work that we've done there. So we added a new column that is gonna give us something that's a bit more intuitive and allow us to rule up if we're having conversations with the supplier, they might not want to know, by default which black ones of this particular product we sold relative to the red ones. Right? So next thing we're gonna do is we're gonna get into chart manipulation, chart configuration. So we're looking at fast situations using drag and drop. We've already seen this. I've built a couple of our channels, which okay. Been a bit boring so far. We're gonna get to something more interesting suit. But, bar charts, we've done this super quickly. We were testing the right code in order to create these. We wouldn't be able to do that that quickly. And therefore we wouldn't be able to have as many iterations that need to get to know data and get to know the tables that we'd be given in as, as quicker time. So ad hoc information from Zueblo, fantastic. If a big engineer gives me a new table from Snowflake or one of our partner products, I can put that table into Tableau. I can do a quick couple of checks, and I can test the validity of the table much quicker than you can in the the interface of the the database I found. So you can explore data through visualizations. So we saw there that the alcohol punch, has four different, under line partners, underline mobiles rather. And we also saw that the Cisco smartphone is the highest selling. These are things that we've done super and we've explored that data, through our visualizations. As we build up our skills with Tableau, we'll build be able to build up our ability to do that even quicker. We can identify insights to answer our questions. So it's something that could be interesting for the business. If we understand their data, then we're more equipped to be able to, answer those questions and also supply pieces of information. Did you know that this is actually the highest selling product? And there's a massive drop off after product five on that table, and we've actually got these five products that are really keeping as a float. Everything else is kind of by the by. So these are kind of insights that we can start to learn once can profile and get the sense of the shape of our data. We can also tell that story using visual analytics. So this is kind of more after we build our chart how do we communicate that to, the the stakeholders that we're trying to communicate that to? Lots that we can talk about in that. We're not gonna too much into the art of visualization storytelling in this session, but, there's lots of other webinars that we've done, that can support that and also the blog I mentioned, previously at block tower, we can go and find out more. So we've talked a little bit about, the concepts behind building charts already. So a bit of a review set here. We've got metadata. Confusing words, kind of think of, you know, Edward Smoden and things like that. So here we've got our data about data. This is all. It's only metadata. Metadata means data about data. The information that we can give somebody includes information about that column that we're looking at in Excel. What does that column include? What information can I give about that column? We've just got to radio things like our data type, for instance. Data types could be a it could be a date, it could be a string, or or a text, It could be a number. That number could have a decimal in it or it could be a percentage. All of these things are metadata. Their information about the underlying data. It could have a geographical role. We're gonna see shortly that we've got a country feel in our data source. The fact that it's called country, and we've talked about it as a geographical field will allow us to plot that in a max while we're alert. There's also default aggregations. So it'll summon sales when I drag that into the view, by default, it gave me the sum. That is a piece of metadata that we've told that field. We want to sum this up. We want to show the total by default. We can always change it later. The default one is considered. Metadata. We could average it out or take the maximum and minimum and so forth. Also, field names are super important. If you're building a data source for somebody else, especially need to have as good an understanding about that data source as you do. So having this well labeled, well named, and that convention is very subjective a lot of the time but generally having a nice naming convention in your data source that is intuitive, and quick to, understand is, is is very much recommended. There's also other descriptive information that we can add default colors, comments, all these sorts of things that are available, in in Tableau. We talked about blue versus green. Again, one thing that you ought to remember from today's session is that there's a big difference between how Tamela treats these two colors, in the in the interface. Lots of other tools also weirdly enough to use blue and green. I don't know if they've stolen it from Tableau or if it's if it's something that's fundamental to the data. Potentially, it just feels intuitive to me now, but I'm probably biased. So blue is our qualitative information. It's information that describes the underlying data. So this could be, the name of the product as long as we lived at before. It could be the city that the product was purchased in. This bucket's up our data, I said grouping before, but it's the same concept. Right? So we want to show all the data that fits into this bucket and all the data that fits into this bucket And that's based on our blue qualitative identifier that's in the rose. It's either or. So things don't fit halfway in between Edinburgh and Glasgow it's gonna be a bad example. What's the things do? But, you know what I mean? In the data, it's either in Edinburgh or or it's in London. It's not somewhere on continuum if that field is blue. That's obviously contrary to the continuous field, which, you know, if it exists on a continuum. He headers, labels, and texts. This is when we drag it onto rows and columns. We will see those headers. We will see our distinct spaces that we have allocated to the, to to the, the fields, values that we're talking about. If we're talking about a horizontal bar chart even for our products, we're looking at different horizontal space that's been allocated to our aqua three hole bridge. Now conversely, we've got the continuous one, and this is where everything's just in this continuing on the scale. So When we look at quantitative data, so this is converse to qualitative data, quantitative data is giving us a a magnitude of value, generally about data that we're looking at in the underlying rules. So, this is the atomic sales that we've looked at could be a number of attendees on webinars. This is, a numerical value often that is gonna be useful to aggregate. Aggregated insight review was talked about different types of aggregation, And there's often a range in between the values, but we want them actually part of visual analytics. We want to highlight that. We want to see that we've got an outlier over here. And we want to be able to make a judgment call and if that outlier is, is is, a significant or not. Exists on a scale and access of potentially even a timeline. We're gonna come back to later on the top of updates. So there is a there's room in between the values that we want to size, that is a green concept. This is a a a chart, a very high level mock up chart where we have a vertical bar chart as we said, something here is on rows. In this case, it's our sales, which is green, giving us that distribution up and down the page that I talked about. And we have the blue objects which are headers, which are discrete and qualitative. Yep. I always get that the wrong way around. And this is giving us that that grouping object that's going across the page. If we were to swap our rows and columns, we would see this split ninety degrees. Now, I think of let's just go back to Tableau here. So I think of everything that we've done in our columns and our row shelf. And if I bring in maybe category over here or let's look at, Let's look, not state. So let's look at market and I drag that onto our column shelf. So here, we've got a little bit more complex here, we're looking at our differences of regions around the globe, and we have assigned basically a, a structure to the chart. We've got a blue object that's going across the page, which is our market, and we've got a product name that's distributed up and down the page. Now these blue objects are giving us basically a structure to our visualization. So we know that there's gonna be a piece of data hopefully in each one of these blue rectangles that I'm drawing and we're cascading them all the way down the page, we could even just drag in our, let's look at category and subcategory potentially down the page here, and then maybe fit that to the whole space. So We have a area that has been directed by the rows and columns shells. We've defined our structure here. Now the green is obviously giving us the distance across this axis instead. And this is basically telling us where a tableau is drawing her marks. Now I've said marks here. We can see down in the bottom left, we've got sixty eight marks. What do we mean by a mark? In this case, a mark is a bar. But that's because we've got some default information over on the left hand side here in this Mark's car. So the next thing after we decide where the mark is being drawn, we want to decide what mark has been drawn on this spot. And this is second key component of Tableau's visualization interface. Here, we could choose that it's going to be a bar or we could say, I want to draw a circle here instead. I could make those circles bigger or smaller by manipulating that size. I could make them a horrible mustard color if I want. So we can define what has been drawn in each of those places using that marks card. So the where is up here and the what is over here. So in this case, I might also want to maybe go rein back a little bit off my mustard circle. So let's go back to nice teal color. And here, I might want to instead of saying all of these marks are gonna be this teal color. I might want to say I want the different columns to be different colors. So APAC's gonna be one color. Amy is gonna be one color. In which case, Market is being distributed across the page. That's defining my vertical distance. I might want to bring market onto color, And now I've got four different colors depending on those markers. Likewise, I might want to put something that's maybe not in the page. Maybe I want to add my, my region. Now I can see there's different regions here, subdivisions of those different markets, and that's dividing up my bar. So I've got another blue object in there. I'm grouping something even further. A little bit confusing, but I wouldn't recommend that. But generally, we can add data to each of these marks cards in order to provide maybe additional grouping or make those, make those properties more dynamic So in this case, again, my market over here, I want to hold control and drag that onto color. This is saying if it's in the APAC, region, then make it blue. I can double click on the legend here, change APAC to be pink instead, and that's then gonna update interview or I could change the whole palette that we're using there if we want. So again, that for me is a little bit about how the Tableau kind of marks card works, so allow me to kind of stick past a little bit of this. What is showing is the mark type. We're gonna come back to that shortly, but this could be bar or a circle or a square or a line even. The color allows us to either have continuous or discrete colors in it. This is the rule that's going to define the color space in our chance. So it could be discrete in which case we have either blue or green or orange or yellow, for instance. It could be continuous. Which case is gonna be somewhere on the spectrum of blue from light blue to dark blue. We can also add detail, and this allows us to aggregate without defining a property why not get there. And then we can also sum up and add some more, information to our tool tip. And the tool tip being the thing that we've just hovered over if we go back to Tableau, this is the tool that appears. So you can customize that in a little bit more detail. So a couple of very hard little things to think about when it comes to building out charts. When we think of the best practice of realizations. We want to choose a chart type that's intuitive. So if we think of, you know, revenue over time, hopefully, we've all gotten our heads on the blind channel or maybe of our chances like that, but we're not looking at a pie chart. Right? So that would be an un intuitive visualization type our level of detail is the kind of granularity at which we're showing the data to our users. Do we need it at that product level or actually we're interested at the, at category level, sort of pulling them all in to see if we're selling more furniture or technology. So that can be very key to the type of analysis we're doing more than the required level of detail of the chart. Data to ink ratio is a bit more cosmetic. We probably want to reduce this, this, Edward Tufte concept, I believe. So the data to ink ratio is kind of about grid lines and, you know, three d charts and all these sorts of things to kind of clutter up the space, try and remove anything that's not communicating the actual data itself be very specific on the way that you want to design that chart. And my go to here is just remove all the different ones that tablet draws and add them back when I need than, like, for instance, if I'm looking, forcing the user to look across the page, then a bit of rule bounding can really be helpful to look across the page. But generally, I use that kind of thing there is bearing on. Hello, in this case, again, think of the best practices, make sure we're not relying on red and green too much because there's color blind deficiencies there. And maybe we want to align that with our brand colors of our organization so we can pull from digital guidelines, or our marketing sort of guidelines in order to get the best color scheme and tablet. We also want to think a bit about the layout, usually putting the most important thing in the top left because that's the first thing people are gonna look at, and that's about when you look at dashboards. Then the user interface, the user journey, when somebody clicks on something, what do they expect to happen? What do you think should happen when somebody, does something in the chat in the in the face, and how can we build up an intuitive tool, essentially, that would be given to face as often as tools to actually get to the answer to not only the first question, but the second and third and question and you choose a manner. So you only have to go one dashboard, I'm lazy, I want to go one dashboard that answers enough questions before I have to split that into the second dashboard. So let's dig back into Tableau and build out some worksheets and charts with, exercise three to five. So The the third tab on the bottom here is building a map. Now this one is quite quick, in order to, in to to configure. Like I mentioned before, we've got these little circular types of data. We can see them like country instance that data type is quite specific in this case, the Tableau, which is a geographical data type. Now Tableau's got a whole database under, underlining, underlying the product of shape files or the the shape of different countries or the positions of different cities. And we can call upon them based on the values that we have in this field So if I double click on country, we can see that Tableau is then gonna assign these countries to the relative place on the map and draw that background image for us automatically. So now we've got a circle in each one of our different countries that we've got underneath the hood, and we can check that Tableos got that right. So Tableau in this case has taken a country like Mongolia, it's taken the value that's in that field and it's matched it against its underlying database. And that allowed us to draw a, say, us tablet down all of this itself, draw a latitude and longitude on columns and rows. Remember what I was saying previously about distribution, longitude is getting distributed across the page because it's kind of the distance around the equator. Essentially. And our magnitude being the distance from the equator in the kind of north or south direction, that's being distributed up and down the page. So we see here that hopefully that's then kind of tying into how capital is creating these visualizations. We've got country in this case on detail because we're aggregating everything up in this case, to that country level. We don't want to define the color by a country we could do, but then we're just gonna get this kind of hundreds of thousands, kind of, view here. So we don't want to do that. We want to keep up on detail and we might want to, for instance, choose to size each of these circles based on the sum of sales. So if I drag some of sales onto the size, Now this is making a little bit more sense, but we can actually get some insights from this analytics. We can see nice big circle in the US because we're making a lot of revenue here, Australia, making a lot of revenue UK, but Papa are doing pretty well as well, I would argue. And then we might want to also look at something that's a bit more, analytics or a little bit more sort of a value judgment on these. So if you're selling lots in France, is that good? Is it bad? If we look at the average profit here, or the sum of profit, even how much money we made and overall, after our cost of sale. If I drag that one to color, Then I can see, orange. Again, it talks about this spectrum of color as previously. If I've got a negative, I can go into orange. If I've got positive. I can go into blue. And Tableau uses a default because it's more color blind friendly than red and green. So our orange ones are the ones we really want to highlight. Turkey's not doing very well here. We're losing quite a lot of money in Turkey. Now it might be a new market, and we kind of put lost leaders out there, and we can kinda have that conversation in our organization. But generally, we might want to kind of do a bit more investigation here, again, what are the top products, what are the things that are driving this value that we're seeing So other things we can do here, I can right click, and I can go into maybe background layers, and I can choose to maybe remove coastlines and to rains and maybe even the state and province borders just to clean that up. That goes back to what I was saying about the data ink ratio. This works a lot better to me than this. Right? So if we just glance at this chart, if we have it nice and clean, remove everything that isn't serving as minimalist, minimalism, And then we want to, then we then we want to kind of remove all of that so that when we look at something, it looks a lot, a lot sort of easy on the eye. I'm even gonna remove that land color, which you can see is just a slight difference in the kind of altitudes in there. It's not relevant to us. The opportunity at this point in our analysis, I'm gonna take it off. Now I've also been asked over here to add a market as a filter. So clicking the x here, going back to our data pane, air drag market onto a filter shell, then what this is gonna do is it's gonna act like a funnel in Excel. Alright. So hopefully that symbol that filter symbol is familiar to people. This is basically gonna prevent the data getting from our underlying data source into our views. For review. So if we have, our market filter in here by drop that in, click on filters here. I can choose which filters I want to add into my visualization. So I could choose to add maybe we just look at EMEA for now. Right? So I can click on okay here, and that's gonna zoom into only things that are in the memory, you're middle eastern Africa. I can now finish this, save it. That's all good. Or I can allow my end user to be able to manipulate that filter. So that's kind of our were funnels and filters may diverge. So I can right click and show that filter, and now my end user can manipulate that to look at maybe APAC instead And I can even say, I want this to be a single value drop down list by clicking on that little go and arrow. Lots of options in here to be able to apply a user interface that my, my end users, like even keeping that all value in here so they can reset the charts. I don't necessarily need my sum of sales, as a size object in here nor am I some of the profits? I'm gonna hide both of those. I'm gonna update my title here double clicking on that title, it's inherited the tab title. So I'm gonna call this my sales, profit. Matt, kind of obvious, or let's just call it servicing profit globally. At that. And then underneath here, it looks like that I used to avoid having these little legends. I like to put in, maybe low profit versus I profit, by country. And then instead of low profit, I'm gonna make that bold, and I'm gonna make that orange. Instead of high profit, I'm gonna make that tablet bold. I'm gonna make that blue. So now very quickly and easily, people looking at this can see, ah, low profit, high profit by country, and I don't then need to have that continuous weird legend of the tablet. So, so that's gonna be useful when I come back to this as a dashboard. Let's look at lighting charts. The lighting charts manipulate, or or require a manipulation of dates. Now dates in Tableau, a very special case. So they are, they are fields over here. So we've got our order date, and we can see in our underlying data, what type of data we're storing So we've got this, at the date level at the moment. We don't have any time information in here, which is fine for now. And we can choose to show this date in multiple different race. We might want to roll this up to the year level to see how we're doing year on year. We might want to look at this on the weekly level. We want to say, however Monday is comparing to Tuesdays. There are lots of different things that we can stipulate about our dates. Now common, common type of length chart that we're most useful, or most used to would be keeping that horizontal, length chart of those teams to the horizontal line chart here, is making us distribute or we need to distribute our order date across the page. So going from left to right, so we need to put our order date onto our row shelf. Now, our row shelf here directly those teams opened up again. Okay. So, our row shelf is now giving us a year on year of the different orders that we have in our underlying data as a structure that we mentioned before. Now Tamara, again, being a bit lazy here. It's just rolled up the highest level, which is our year. If I right click on this, I want to change and manipulate the type of data that we're showing the granularity of that data. And we can see but halfway down the page here, we've got two different distinct options that we can use. So fixed value, these look very similar. Right? So we've got year quarter month day and more. We can look into that. We've got year quarter month, week number and day. So very, very similar here. The second one that we've got here is the date value. Tableau often uses this type of terminology And the date value is gonna allow us to see trends in our data. Trends in our data, are gonna be highlighted based on the fact that if you notice, we've the year in that little example. So we've got May twenty fifteen compared to just May. So this is gonna allow us to show trends, and it's gonna give us a timeline these are the two terms that I would kind of associate with our date value. And conversely to this, we've got our date part. And in this case, we are bucketing up our data. The reason I've done this in blue is not coincidence. We are grouping up our data based on these, date parts that we're, we're highlighting to our end user. So in this case, we're looking at this in buckets without next. And the buckets are going to be grouping our data based on a snippet of the data, part of the data. If we take, for instance, let's change to a more neutral color here. If we take the date that's given in the example of the eighth, of May, two thousand and fifteen, then we can roll this up and down in our trend line to say, give me it at the day level. I want to see an axis with lots of little dots on it, and that's gonna give me the day level, of our of our date. So if we've got a year's worth of data, then if you see three sixty five different inflection points on the timeline, or we could rule it up to the month level, in which case we're gonna see twelve inflection points on the timeline, one for January, one for February, But again, this is gonna be different. If we've got multiple years, if we've got two years of data, we want to see twenty four different points on the timeline map. And then we could just roll it up to that year level. If we want to look at the leading part, then we might want to look at for instance, the day of the month. In which case, we're literally just scripting out this value eight. We're taking that value and we're gonna maybe plot that as we've got our years here instead. So we're gonna get one to thirty one, and we're gonna see a dip when it gets to after the twenty eight probably. Because not all months have thirty one days in them, but we will still see that value because it's such a distinct value. We could also have a look at all of our January together all of our February together by stripping out that lock part, or we could strip off that year part instead. So that's a little bit about the difference between the two. You'll get the wrong one. If you're like me, you'll select the wrong one, but understanding that difference will allow you to then fix the problem before it becomes a bit more of an issue. In this case, I want to look at the month, but I want to look at it as a timeline over all of the years in my underlying data, the four years that I've got. So I'm right clicking gonna go down to our month timeline here, which is a date value. Clicking on that, I can now see March twenty seventeen. November nineteen, that looks good. So I'm gonna drag my sales onto rows to distribute, up and down the page, giving me that time axis and I can see that I've got this nice seasonal view going up the page. If I was to change that to be the other month, but We then see this reduces down to only twelve marks, and I would probably use that as a bar chart instead. I review that as a bar to make sure that we don't think that one's leading into the other because they might not necessarily be. So timeline sales over time. We've got our month of order date in there. I think we're good to go. So let's now look at a very quick bar chart so that we've got this for our analysis later when we build out our dashboard. Now this bar chart is gonna be a horizontal bar chart where we look at our sales category and our sales subcategory. And we want to look at the sales figure over those. So I'm gonna actually just cheat here. I'm gonna select category select subcategory holding control so I can multi select them. And I'm also gonna select sales here. I'm gonna go into the show me pain over here, which is definitely cheating, but in this case, allow it. And I'm gonna click on one of these pre built, versions of charts over on the right hand side. So if you're looking for a bit of inspiration, this will build some charts for you. By configuring this page. And at this point, it looks quite good. It looks like it's done pretty much what I wanted it to do. We've got sales, a category, a subcategory, split out. And now I want to also add my profit onto my color shelf, and then I'm gonna use this fitting panel here say I want to actually make sure that this fits the entire page. This is to be used with caution if we're using even further since our product that we looked at before, because it's gonna smoosh everything into the view. We might sometimes want that probably not when it comes to product names, in this case, we don't want to have any school bars because if we do get any more subcategories next year and the year after, we don't think it's gonna compress the view for, to the point where be visible. If you've got something very granular, you probably want to have the school wire instead, in which case you don't want to fit the entire view. So I'm now gonna sort that descending using my little sort objects up at the top here. And, now I'm happy with how that looks subcategory sort. I can see that tables, bit of a joke from Tableau kind of a a hard, a hardy hard about the fact that tables are losing money. We shouldn't be doing tables in of visualization worlds, we should be building visual charts and, end bar charts and that sort of thing instead. So you always see that cables is losing a little bit of cash as a bevan Easter egg there. So now I've got my bar chart. Everything looks good. I could choose, to maybe show the sales on my label here, and then I can right click and get rid of the header. I want to do that. Despite dragging with Google sales onto my label, then right clicking that header and gateway. And does that give me everything I need? It kinda does. This looks good. I might want to format this by right clicking and like I said before, essentially getting rid of those grid lines because they're not really serving me anymore, and I can make a choice on my, horizontal lines. They do split up different categories here, so I'm gonna leave it. Now we get on to calculated fields, how many times? Thirteen minutes pass. Okay. So calculated fields here, I'm wanting to create a metric for my sum of profits divided by my sum of sales. So just because we've made I don't know five hundred k of profit in India if that's, because we've actually sold, a billion dollars a pound's worth of stuff. That's absolutely not a great, and not a great sort of margin that we've made on that. Right? So, we need to know that kind of margin calculation. We want to then essentially average out that margin over our different, our different countries to see, exactly which, which country has sold the most, and which one has made the most profit out of the sort of the sales that is made. So in order to get this, we don't actually have this information in our data currently. We don't have the margin on each sale. We don't have the margin for each country, and this doesn't exist. So if I were to create, let's just again go back to our, our visualization that we made before and add profit in there, So I've got very quick view here of our categories, and we've got the, sales and profit. Not segment subcategory. There we go. So we've got cables doing badly again. Now I want to add in here my profit margin as a percent So the way I'm gonna do that is I need to add another column into my underlying data. Now I'm gonna create a calculated field here to add that column and insert the syntax similar to adding another sorry, some syntax in Excel. So I could either, use the little down arrow in the data pane to create a calculated field. I could right click in data pane, create a calculated field from here, I could go into the analysis menu and create the calculated field from here. I'll do the exact same thing. So I'm gonna create that calculated field. And we're gonna see this nice helpful little window appearing, that I can add my syntax to. The first piece of syntax I need to add is the name of the calculated field. And I'm gonna call it profits ratio as instructed in the caption. So now I can zoom in here and I can a look on this right hand side or helper functions here. If you want to use any of the plethora of different functions tablets got available to it, then you're more than welcome in here. And if we want to look at ceiling. It'll give us some more information about exactly what ceiling is gonna do with an example. I don't want to use any of those at the moment. What I'm gonna do is I just want to add up all of my profit. So I can start typing here by starting to write sum. We can see that sum is a very common function in Tableau. And I can hit tab and it's gonna give me the curly brackets that it needs. Then I'm gonna start taking profit. Now it's gonna show me previous value because that's a function. It's gonna show me products, make sure me product name. I don't want any of them. I'm gonna pop it here so I can hit tab again, and it's gonna also fill that. Notice that it's it's orange. That's because it's a field our underlying data. Then constant mathematical functions apply here, I want to divide divide that by my sum of sales and what I can do is some of the sales already is this here, so I can just pick that up and drag it in nice and quick. So I've got my summit profit evaluated by my summit sales. I can click okay here, and we're then gonna see this new calculated field appear in my list of measures. So now profit ratios over here on the left hand side, can alter the metadata even further. So this is likely gonna be a percentage. Right? We're gonna get somewhere between, you know, if we're making a hundred percent profit because for some reason it costs us anything or it costs us tap as much than that later. But, if we're getting one hundred percent profit margin, then we want to show that as a percentage. So I'm gonna right click here, go down to default number format. I did that too quickly. Default properties number format. And I want to make sure that this is a percentage. You're gonna keep one decimal place in here. Now when I bring this into the view, I'm gonna see I've got my profit ratio over here. So I've got minus eight percent eight point five percent made on my, my tables, and I've got nice healthy profit margins made on each of my other subcategories, too. So I've created a new piece of data from that underlying data and you might see the advanced data preparation in action when we're moving through in our analytical workflow need some more data. I don't want to go back to the database on it. Just want to create this in the visualization layer. So finally, before we get to the dashboard, I'm gonna create the very quick and easy, segment, KPI. So I want to look at each of the segments across my data source and I want to bring in some information ahead of level. So I want to create that kind of banner at the top of my dashboard that says this segment's made millions and this segment hasn't made millions. So that gives a very high level of pieces of information for my user. Gonna do this quickly. So I'm conscious of time here. I'm gonna copy this text. I'm gonna bring in the data that I need for it. So I need to bring sales onto detailer. As she sails onto text, I'm gonna bring segment onto my column shelf because I want that to distribute over the page. And I want to bring profit in there as well onto text and profit ratio onto text as well. Now these don't look particularly beautiful. So this is where a little bit of, behind the head out here, and bring my segment onto x two. Is where I need to kind of make sure that this chart itself looks a little bit cleaner. I want to hierarchicalize if that's a word it is now. I want to change this in my text editor on the left hand side here. And now I've got this kind of rich text editor that's kind of done a bit of a kind of mail merge type thing where it's got these soft coded, values in the year. I want to basically manipulate this, this style. I'm actually just gonna copy that style in so I have it on my clipboard still. And then I'm just gonna insert that sum of sales. I'm gonna take segment, and I'm gonna pop my segment in there. It hasn't even marked as it. Let's go back down to adding that size ten, and then put profit in there. That's better. And then proper ratio in there. So now I can see that's working quite nicely. Might want to roll that sales up to maybe a hundreds of thousands, so I can right click on it look at the pain level number currency. Let's go in pounds in UK here. Let's look at That's probably enough. Yeah. So let's look at it in k. So we're gonna filter this down later. So once we look at this in k, we can add obviously more rich text here to highlight the that is a, is a number or a currency. Let's go standard UK. Let's take away the decimal places. We could even add the word proper in there to make it obvious, but I think that's okay now. It's very quickly we've just pulled together our head of level segments. So now we want to get into creating a dashboard from the visualizations that we previously created in this last ten minutes here. You could use this exercise sheet here. You can hold alt and just remove that, exercise caption if you're coming back to this later. What I'm gonna do here is actually I might just start a brand new one for simplicity. So if I go into this new dashboard button, want to create a new sheet, it's this one to come to create a new dashboard. It's this one. So I'm gonna create a brand new dashboard so you can see exactly what this looks like by default. Now first thing is to choose a size to this dashboard. I like to fix this maybe a little bit bigger than Tapl's giving me by default, maybe twelve hundred by eight hundred. This is giving me the size of the canvas that I've got to play with here. I'm gonna give it a nice intuitive name like global sales. I've got that map that I'm gonna lean on heavily here. So I want my map to be front center, make it global, make it regional, make it have something that indicates our, that are sort of geographical information. Then third thing I'm gonna do is I'm gonna show that dashboard title in my view, which is gonna add a text object. So this text object's being moved to the top here, gonna then double click it. I'm gonna make this nice and big, bold and middled. I can perform some string manipulation up here. So now I've got my global sales as a header up here. This gray outline shows me that I've got an object on my canvas. This is what the gray outline is giving me. So this object could be any of the objects in the bottom section here. Most common ones are gonna be text. Or actual worksheet itself. So you'll notice here that on the left hand side where we previously had a data pane in the dashboards, setting, we've actually got the worksheets that we can then bring in. So previously, the components of our worksheet are data, and the components of a dashboard are chart So if I want to bring in a chart, I can do if I want to change that chart. So I want the colors to be different or I want the roles to be on columns and columns to be on rows. If some fundamental change, I have to go back into the worksheet and I have to change it there. So we're creating all those component parts, and then we're bringing them together in the visualization layer itself, here. So I now want to bring in the most important piece of my global analysis, which is gonna be my map. So I'm gonna pick up the map on the left hand side. I'm gonna drag it into my view. And wherever we see this gray, we can kind of position that wherever, wherever we want. So our map is looking good here. We might want to make some tweaks to this later on, but let's keep going for now. So I now want to bring in maybe that timeline. So the timeline's pretty fundamental. How are we doing over time in terms revenue. I'm gonna bring that line chart in. Maybe I could drop it drop it down here. It feels maybe a little bit intuitive to have a timeline at the bottom. You'd expect that at the top, I suppose, but I'm not sure why I'm basing that on. So let's just change the type of font here to be tam below the twelve. I would say make fonts nice and consistent, even within the Tableau spectrum if you are using Tableau fonts. If you have your headers being Tableau twelve on each chart, then make sure they're all Tableau twelve in each chart. So going over here, changing that to be bold. So now we've got our two visualizations is working really well. The last piece that I want in here is to have my category and subcategory on the right hand side. I'm gonna bring in that bar chart that we created earlier. Is that the right one? Yeah. I think we've also got our KPIs. Let's drag them at the top here. Give that enough space So you can see it's struggling to fit all of that information in. So I'm gonna go into my underlying sheet here. I'm gonna make the sizes smaller. I've been a bit ambitious here with their sizes. So that hopefully will now fit. Over here. Yeah. Perfect. Okay. So now we've got our KPIs for consumer corporate and home office. Fine. And we have our map. And now I want to bring in my bar chart here. So with this version, we can now sort of tile this approach. So we can have anything that's in the view. So one trick I've got here is that I have what's in the view, and then maybe I want to bring this over here so that I have more space available, to the chart don't need that profit because we've got this legend. We can choose if we want that or not, and actually I can choose to have this market, filter I can choose to float that and then get rid of that underlying object to give us a bit more space. So this floating is kind of more similar to kind of power point layout. We can choose where this object sits in terms of pixels. A little bit of tidy up here. Pick that size twelve involved again wasn't it. Good. And yeah, we can always manipulate these charts a little bit better. Does that need as much space as it's got? Yes, it does. Does that? No, it doesn't. Okay. I've also got some sort of header there. Yeah, there we go. Okay. So Now, we've got our component pieces of rapport different visualizations in our chart. We might want to allow for flexibility or people to be able to answer questions on this. We've answered a couple of questions so far, but maybe we want her market filter to apply to everything. So we're gonna click on apply to worksheets, rather than just using the map, which is the one we started it on, I want this to apply to everything. So we can see we've already built an interface here is gonna be valuable for people across different markets to get the sense of their head of level information. Now if we want to look at Turkey, we want to drill into that then we can do so by clicking on the funnel here, the users filter funnel. If we click on Turkey, now I can see exactly what's going on in the Turkish area in terms of the timeline, in terms of the different categories. Everything's losing some money here. There's some sort of big hairs. We might want to maybe go in and look at adding profit ratio there instead, something like that. So this also works with click and drag. I've got kind of custom regions. Maybe I want to look at the south of the African continent. We can see what's doing well there and what's doing poorly. So really nice quick analysis to be able to answer. Your own questions across the globe in terms of those transaction sales. Likewise, maybe I want to perform the same thing on a timeline level. So if I select time, if I select a particular per month, how does that differ to other months? Oh, this month, we've made a lot of sales in India for instance. That can be very useful. And, obviously, we can do the same on our, categories. So let's look at tables. Where's really bad US? Really wanted losing my own tables as well as Indonesia may want to look at suppliers for tables in those areas. So as I promised before, I'm going to, sign into my, abroad server. I have time. Let me try and get into it quickly. And I am going to sign in here by clicking on the server button in the top. Just showing, removing that from screen so they don't see any credentials, that kind of thing. So now that I'm all signed in, I've got my tablet server up here. I want to push this so that my other colleagues at Zoltan can see it. I'm gonna publish this workbook onto my Tableau, instance, and I'm gonna push that into Tableau examples. I'm gonna call this or Tableau with some exclamation marks. I'm gonna remove everything that's not a dashboard, but I'm only gonna show that global sales dashboard. I'm gonna choose to, some extra configuration things and here permissions are fine and better data source is fine. So I'm gonna click on publish and that's then gonna upload to this chart definition into my, Tableau, Cloe, Tableau certos, so I think in this case, but again, it looks very similar. It's gonna open up in my default browser. And I can see a little bit of the Tableau cloud or Tableau server instance here, and I can have a look at this in the browser, and this is what my colleagues would see if I send them this link. I can then share them. This tool that I've created for them in to explore their, their sales figures for, the different regions. Every piece of functionality that I've created in the fundamental desktop tool that we've got here is available. So I can filter this down, and I can combine different filters here to have a look at and unique points within that map to see if these are all combined together. So again, we have touched on a very high level, perspective those a good view of Tableau. There's a hell of a lot more to it than that. But hopefully that gives you a bit of a taste out of, of how Tableau works. We've got some more content here that we actually got through throughout the session. So we're looking at calculated fields. And these can get complicated build dashboards, we're displaying lots of data, and we've done all of our exercises before sharing, distributing our work