Tableau 2020.4 new feature Livestream

0:25Hello.

0:28I'm just setting up, we're just getting started, and we're going to be covering what's new in Tableau 2020.

0:354.

0:35I can also hear myself somewhere, so I've got some feedback at the moment.

0:39Um, but yeah, uh welcome to this live stream, everyone who's joining in.

0:42I assume we do have some people watching on

0:44the stream.

0:45Um uh this is gonna be a long session.

0:47Um do that from here.

0:49Siri seems to want to join in, so I'll be taking my watch off.

0:52Um it's gonna be a long session.

0:54We're gonna try and cover all the new features in 2020.

0:574 um in one go.

0:58Uh previously we actually did them as separate uh live streams.

1:02But I think that sort of uh strung it out.

1:03And in this particular release, there's actually some features that have a lot of synergy.

1:07So I thought it would be great to just show everything working together seamlessly in one go.

1:12So we're gonna be uh getting stuck into that.

1:14uh in a moment.

1:15Um if you're just watching, if you're just joining, uh please leave a comment.

1:19I can see the comments myself, so um please uh make sure you leave a comment um in the

1:25chat section of the YouTube interface.

1:28It's normally on the right hand side depending on how big the video window is, but if not you can just enable it and it will pop up.

1:35depending on what app you're using.

1:37If you're on mobile, then you might have to scroll down.

1:39If you're on other devices it might not show up.

1:41So if you can engage that would be great.

1:43I will be paying attention to my Twitter as well.

1:45So

1:46If you've got no other way of reaching out, you can absolutely just send a chat there.

1:51Got a few people joining in.

1:52We've had a message from Mario, Chital as well, and Dimitri.

1:56Dimitri was actually in the last live stream.

1:58I remember Dimitri

1:59Dmitri, so hello Dimitri, how are you doing?

2:01Um, we're gonna get started in about two minutes.

2:03I'll give people a bit of time to join in because not everyone's always punctual.

2:08But in the meantime we'll just we'll just sort of um warm up as it were.

2:12I'm just gonna make sure everything's working technically.

2:14Um if you can hear me please let me know in the comments.

2:16It's one of those weird things where you you can't hear yourself.

2:20easily sometimes when you're doing sound checks so I think I sound okay but if you can't hear me let me know otherwise just give me a thumbs up or yes that's okay in the chat and then I'll know we're good and we can get started

2:31Um while that happens I'm just gonna change a few things on my end.

2:34If you see me looking to the left over here, it's actually because I've got um I've actually got the

2:39live stream setup here so I keep looking over to my left hand side just to see what's going on but essentially that's what I'm sort of trying to manage then I also have a chat window here so if you look see me looking to the left it's not because I'm bored

2:51just because I'm trying to manage several things at the same time.

2:54So hopefully that's all okay.

2:56I will also be getting messages from colleagues if I'm absolutely messing up.

3:00Thank you, Steph, for the message on Convoy.

3:03So yes, I've got lots of places to be paying attention to.

3:05But for the next hour and a half, the Q focus is 2020.

3:094.

3:10So we're going to get stuck in.

3:12Before we get started, I always like to highlight that we do have a whole series of events.

3:16We have lots of great content on our YouTube channel as well.

3:19Let me start off with a YouTube channel actually because

3:21Of course I'm the I'm the YouTube um fan here.

3:24So let's go to the um uh YouTube channel here.

3:27We have lots of lots of great content over the last uh year or so, ever ever since lockdown really sort of forced us all

3:33to become more digital and uh spend more time online.

3:36We've spent loads of effort and time adding really good quality content uh onto our YouTube channel.

3:42So absolutely go check it out.

3:43We're just shy of 10,000 subs.

3:45So I'd love to get to 10,000 subs before the end of January.

3:49That would really help us out.

3:50um help uh spread the message out.

3:52We've had some absolutely fantastic content from the data schoolers at the information lab.

3:57They've produced uh some amazing series um in the channel.

4:00You might not even know about this

4:01actually if you haven't watched our channel before.

4:04Um there's a whole section on Tableau Prep.

4:06In fact if I just go to all videos you can just see sort of the consistency.

4:10Pretty much every week we are putting out a video showing you how to do really simple but also really complicated things.

4:16in tableau so there's a whole host of content here I bet you if there's something you're stuck on then it's going to be covered by one of these great uh individuals

4:24of the data score and it's going to be a video that is normally no longer than five to ten minutes.

4:28The longer ones are obviously live streams or webinars, but it's a really easy way to get involved.

4:33The other place to check out is of course our meetup page.

4:36So this is the meetup page for this livestream.

4:39And essentially this is where you can come to find all the events that we run.

4:43up front.

4:43Um short of you know engaging with us on social media and finding out there, you can actually see a whole calendar of events that we've got lined up.

4:50So we have some training sessions every week.

4:52That you can take part in.

4:53Um, but also we have some slightly off the cuff sessions based on you know what our data schoolers are seeing in the field, what our consultants are seeing in the

5:00the field and of course uh live stream like today's where we're touching on new features so whole range of content you can always look up front um ahead of time to see what's coming on but also just you know pay attention to uh social media when we talk

5:12tweet these events out and um you can find out that way.

5:15So that's pretty much the ways to get in touch.

5:17It's always good to cover that at the beginning so that we can focus the entire session on 2020.

5:224

5:23Alright, to get into 2020.

5:244, what I'm gonna do is I'm actually gonna go on to this page here.

5:27So um Tableau this year have actually released a very nice video.

5:31I'm not going to play a video on stream.

5:32That's not what we're here to do.

5:34It's a six minute video and it covers all the features.

5:36So this is a great page.

5:37I think this is the first time they've done this.

5:39I'd like to think they've done this because

5:40uh the the YouTube content has been going through the roof so that they're sort of putting their own uh best foot forward and creating summarized content for us to

5:48get involved with it's a video that you can download put it in your own coe and it covers all the features in in six minutes i actually used to do this myself in ten minutes they've done it in in in less

5:57So that's great to see uh this this time round.

6:00And then the main page that covers all the features is of course this one here.

6:04So I'll go back to the top.

6:05This is sort of the highlight page for every release.

6:08It's probably the best page to send to people uh as a link.

6:10It's the easiest way to send

6:12of make sure that people can see all the features in one place and um it's nicely put together they sort of cover the headline features at the top they give some quotes and they they sort of uh really do their best job to show off these features

6:24and typically using GIFs videos if necessary and but they have this great filter at the bottom and it this filter always surprises me because every year it gets longer.

6:33Every time you hit this drop down it gets a little bit longer and the list of products

6:37gets longer and longer and longer.

6:39Um I know there's a couple of products missing in here.

6:41For example, we don't have Tableau CRM, which is a new addition to the Tableau family because of the Salesforce.

6:47acquisition but it's all going to be uh here essentially as the tableau architecture and the platform gets more and more encompassing and we get more and more capabilities.

6:56But this is the roster of features that we have for 2020.

6:594.

6:59We have a whole load of features.

7:01And before I get stuck in, um this was actually the first release where I was actually really confused.

7:07sort of what to tell people because I spent most of my time in the browser.

7:11I basically managed to use 90 to 95% of these

7:15features entirely in the browser as well as doing things that I couldn't do in the browser before entirely in the browser.

7:21So I'm going to try and mix that up a bit today.

7:23You'll notice that I talked about authoring and not desktop because I think that's actually going to be the way to talk about building workbooks and dashboards in the future.

7:31You'll be an author.

7:32Where you do it won't really matter, whether it's on desktop or in the browser, it's all going to be the same experience.

7:37The other thing to bear in mind is that obviously some of these are slightly difficult to show on stream.

7:42So things like the data connectors I'm not really going to cover and because all data connectors are

7:46are more or less the same.

7:47They all have their own quirks and there's hundreds of data connectors out there for Tableau.

7:52But I will touch on some of the nuances around these connectors.

7:54So I'm not going to skip past them.

7:55I'm going to highlight why they're good if there's a little bit of context as to how they're built or this

8:00something new about them.

8:01Okay.

8:02So in order to do this, I'm going to try and balance my time between Tableau Online and Tableau Desktop.

8:07So to start off with, I'm actually going to go over to Tableau Prep and we're going to start in the browser

8:12So let's go over to Tableau Online.

8:14Now this is my own Tableau Online instance.

8:16It's actually a beta instance.

8:18It's used for testing purposes.

8:19But you can see here that it's actually currently on Tableau 2020

8:24So this is the most recent release.

8:26If you ever wondered what this 10AX is, that's essentially the cluster, the AWS cluster.

8:32that your Tableau Online instance is on, depending on where you sign up to Tableau Online, you'll be in different locations.

8:39So if I switch to one of my other ones, for example, if I switch to

8:42the information lab uh demo instance you'll see that this has uh du b one in front of it.

8:48It's actually gonna ask me to log in.

8:49I should already be logged in

8:50There we go.

8:51And if I go over here to the about online, you'll see that this says dub01.

8:56So this is a really really good place to uh you know

9:00Try the best of Tableau in my opinion because they are really putting a lot of effort into making sure that the online experience is really good.

9:07And so that's what we're going to be using.

9:09mostly today.

9:10In the event there is a bug.

9:11There have been some bugs in the browser.

9:13So if then in the event there's a bug, I will switch over to desktop and we'll show as many features there.

9:17If there's something different in desktop, I will also show that.

9:21But pretty much everything's like

9:23like uh if that makes sense.

9:24So uh let's get stuck in.

9:25I'm actually going to switch back to my uh test instance because I actually know that everything worked there really really well.

9:31We'll come back to the um

9:33uh information lab demo instance actually because I've got something a little special to show you.

9:37This is actually something that's out in the public.

9:39Uh Tableau are asking people to try it out and give them feedback.

9:42So

9:43I'm gonna kind of leave it as a mystery of what that is, but uh if you stay till the end of the stream, I'll be showing you something that's only just gone into preview yesterday.

9:50So it wasn't part of the beta, it went into preview yesterday

9:53and I'm going to be pretty excited to show you how that works.

9:56Okay, so let's get stuck in.

9:57I'm going to go over to explore here and this essentially just shows me all the content on Explore

10:02my Tableau Online instance.

10:03Now if I hit this drop down arrow at the top I able to see all the different types of content on my Tableau Online instance.

10:10If I'm on Tableau Server, it's more or less exactly the same.

10:13And what I'd like to do is I'd like to open up a data source.

10:16And you can see here that I have a superstore data source and I also have a data source called

10:19prep in the browser.

10:21I actually created this a while back when I was trying to create demos for this product.

10:25But we're not going to touch any of those.

10:26What I wanted to show you is that I have a superstore data source already here, but I'm actually going to connect to a fresh Excel file

10:32I'm going to upload that here onto Tableau Online.

10:34So I'm going to hit new and the first thing I'm going to do is hit flow.

10:37So what will happen is it will open up a new browser.

10:41And essentially we get straight into it.

10:43So if you've ever used Tableau Prep, you'll notice that actually on desktop it can be a bit of a performance, uh, you know, hungry application, but on the browser, um at least here in Tableau Online where Tableau are looking after the resources.

10:55somewhere in the cloud, it's very, very smooth.

10:58So um let's just go ahead and connect to a data source.

11:01I've got some files on my desktop.

11:02I'm not going to connect to a database.

11:04That's sort of asking for trouble on a live demo.

11:07But I'll go ahead and connect to an Excel file.

11:09Actually go ahead and connect to a file on my desktop.

11:12So I have some superstore.

11:14Sales Excel file.

11:15You've seen this before.

11:16It's nothing new.

11:17It's nothing different to the one that's in the Tableau repository.

11:20It's a three-meg file.

11:21It will upload very, very quickly.

11:22And what's great about this is that when Tableau actually uploads the file, it's actually keeping the Excel file in

11:28in its original format.

11:29It's not changing it to an extract.

11:30So you're actually working with the original asset.

11:33You can see that here on the left hand side because I get all the things I'm used to in

11:37in in the prep for desktop so I can see the orders, people's returns, and everything works uh as it normally does.

11:43If you're wondering why I'm seeing order twice, that's because these little icons

11:47Refer to a specific feature in Excel.

11:49The orders is essentially just the sheet as we see it in Excel.

11:53But this one at the top here, it's actually a named range in Excel.

11:56So named ranges come through as their own table.

11:58So if you ever wanted a hack

11:59as to how to clean up an Excel file very quickly just create a name range on the section of the data that you need and that will come in on its own in Tableau.

12:07So let's go ahead and drag just the sheet that we want in Tableau.

12:10And as you can see it's really smooth.

12:12I can

12:12just drag it in, add uh view and clean step here, and it will just works as as expected pretty pretty much.

12:19Um some context behind this Tableau Pro was actually built uh as a browser first

12:24and then it was packaged for desktop.

12:26So most of what we've been using is actually just a desktop application and it's been very very easy to to um

12:34Just spare me one second.

12:35I think I'm getting a message from someone.

12:37It's been very, very easy to use.

12:39If I just let me just set this up so I can actually see this a little bit easier.

12:43One second.

12:45Uh I feel like this is an important message, so I'm just gonna make sure that I can see this cool.

12:52So

12:53We've just got a standard um uh prep uh instance here.

12:57Now what I'm going to do is I'm going to split this as you normally do in Tableau uh prep on the desktop.

13:03I'm going to do an automatic split just so you can see that working really really quickly.

13:07In prep, what it would do is actually create three new columns.

13:09So we've got our original order ID column here, and then we've got the three new columns available.

13:14on the left hand side.

13:15Now the other thing you can do is you can of course bring in other sets of data.

13:20So of course I can connect to other things, join them cloud.

13:22clean them and bring the data together, everything as you'd expect.

13:25Now, there's a few quirks with working in the browser and it's probably worth highlighting this before I build out a flow very, very quickly.

13:31quickly okay the first thing is if I just go ahead and close the tab I didn't save anything okay I didn't hit save anywhere and you're probably wondering well where did it go?

13:39Well all your work is automatically saved in the background let me just show you where that is if I go to my home

13:45I go to explore and I go to default, you'll see that I have this area here.

13:50And you can see that I've got two.

13:51I've got one which I worked on the 8th of December of last year.

13:54last year.

13:54So this draft here was saved on the 8th of December.

13:57So this gives you a context of how long it's keeping these drafts for.

14:01It never gets rid of them essentially.

14:03Whereas this new flow here, I never named it, I never published it, I never did anything to it, created just a few minutes ago, um 6th of Jan 613.

14:12So everything that you do as you're doing it is being automatically

14:15So if your computer crashes or something goes wrong, you can absolutely just very quickly come back here to the default page and hop in and everything is

14:24good to go.

14:25We carry on exactly where we left.

14:27If I click on the clean step, it's almost like we didn't miss a beat.

14:31What you then do have to do before you run any flow.

14:34is of course publish it and it's this is different isn't it because on desktop we're used to saving things but here on prep in the browser we actually have to publish

14:42things so let's go ahead and publish this it will ask you where to publish it's a bit of a weird terminology i'd prefer that to say save because if i'm working in the browser

14:50Publish seems like the wrong word if that makes sense, because publish used to be from when you're going from desktop to server.

14:56So in essence, if I'm in the browser, it should just be called

14:59save but I'll save it to my default folder I'll call it um 2020.

15:034 demo and we'll um I think I put a dot there you're not supposed to um actually no I'm supposed to um choose somewhere to

15:12save it as well 2020.

15:144 demo hit publish I've put it into my default space and now that's basically saved or published in in this

15:21particular instance I get a success message at the top and now if I add a very simple output here you can see that this is pretty much ready to go I get the standard output options.

15:30Okay.

15:31Now, the other thing to bear in mind is that of course Tableau Prep needs to be able to run the flow.

15:37Okay.

15:37And in order to run a flow on Tableau Online,

15:40or Tableau server.

15:41There's a little catch here.

15:42You're actually going to need something called the data management add-on, which is an additional add-on in terms of licensing for Tableau Prep and

15:49Tableau prep conductor.

15:51And so it's a little bit of a convoluted tree of uh sort of relationships, but you've got Tableau Server and Tableau Online, they're the core product, and then there's something called the data management add-on, which essentially is two things.

16:02It's um

16:03It's part of the d it's it's two things.

16:05It's the data catalogue, which is essentially linked to the Tableau lineage features that you get when you look at um tableau online assets essentially on Tableau Server Assets.

16:14Then you've got Tableau Prep Conductor, which allows Tableau Prep to be automated on Tableau Online and Tableau Server.

16:20So those things are sort of all bundled into one product which is called the data management.

16:24add-on so you do need that because in order to run the flow you actually have to have that enabled so when I hit run here um I'll show you a little quirk with this in a second in order to do that in order to have that running on the server you do need that data management

16:37add-on otherwise you can absolutely uh you know publish a flow but you won't be able to sort of automate anything you'll you just sort of be sort of s using table as like a nice place to save the files if that makes

16:48Makes sense.

16:49So the other thing I'm going to do is I'm going to set this up.

16:52Notice that on the output page we don't have any ability to export to a CSV.

16:56They've removed that option, whereas on desktop you'd have that option.

16:59So bear that in mind if you're going to be working in the browser.

17:02You have to publish as a data source or to a database.

17:05To be honest, the database is the most important thing there.

17:07And most people shouldn't really be writing, you know, flat files to their desktop, really.

17:11Doesn't

17:12sort of make uh sense but if you have to you have to and that that's fine I'm gonna do a published data source to keep this simple I'm gonna send it to my default folder as well I'm gonna call this uh 2020 uh for demo output

17:26Great, and uh I'll just call this uh done live on stream.

17:31Okay, and we're gonna create a table because it's a brand new uh item and we're just gonna hit run.

17:37Now the thing is when you run, it's not like desktop, you have to publish your changes.

17:42Because I've changed something here.

17:43It's asking me if I want to publish the changes.

17:45So I wish it would do that automatically because now that I've done that

17:49You can see it's successful.

17:50I then have to hit run again and now the flow is actually added to the queue.

17:55It's not going to run in front of me.

17:57What I'm seeing right now is the preview.

17:59When you hit run on tablet prep in the browser, it gets added to a queue.

18:02Let me show you where that queue is.

18:04If I go back to default here, uh, this is the table online view.

18:07And I go back down to this option here on the bottom left called jobs.

18:11When I click on that, I actually have this job pending here.

18:14And you can see that it's joined a queue.

18:17Essentially, I'm actually in the global queue for all Tableau Online users

18:21in my Pacific Tableau Online instance, which is 10AX.

18:25I always forget where AX is, but I think it's Arizona, one of those AWS.

18:30locations but essentially it's a pending item so the run of that flow the full run of that flow goes into a queue however

18:38My preview happens immediately.

18:41So the reason this is important is because if your preview is, let's say, uh limited to the first million rows and you've only got

18:4710,000 rows and you're fine because you're seeing your output right here.

18:50But if your data set is really large, let's say you're connected to a database and you're only sampling your data, which Tableau Prep likes to do, then you're not going to get your output straight away like you would on desktop

19:00where you can see a progress bind run the output straight away.

19:03If I go back here and I just hit refresh, I think this should actually be done now.

19:07It gives you

19:08to uh time so it gives you a run time it's still pending actually so it's not run um I assume this is busy because um the US is online at the moment and that seems to be the sort of the more busy um

19:20table online instance as it's were but it's been queuing for 1.

19:233 um minutes if that makes sense so we're gonna wait for this um to

19:28To finish.

19:28Um, if I go back over here, I'm gonna actually make a few more changes and we can come back and look at that uh sort of job and see how long it took.

19:35So I'll close these two.

19:37And I'm gonna go ahead and do a few more things.

19:39I'm gonna add some more items from this superstore excel.

19:43So we've got people file here.

19:44I'm gonna work pretty quickly here now because I'm running a bit behind.

19:47So I'm gonna work pretty quickly here to add this in

19:50And this IR person file is actually just going to give us the names of the managers.

19:55So what I'm going to do is I'm going to join it on.

19:58I'm going to remove my output there and I'm going to try and make sure I

20:02I'd do a left join by clicking the left hand part of the circle and the inner part of the circle as well.

20:09So that brings through all 9994 records, which

20:12Which is good, although I do seem to think I had more.

20:15So I had 10,000 rows here.

20:17No, 9994.

20:18It's rounding that up.

20:19So that's correct.

20:20We've got the right number of rows coming.

20:22through and then I've got returns now my returns data has this is a little bit of a weird one because um some of these some of these have multiple rows um

20:33for the same amount of data.

20:34So if I if I just actually this is a this weird change they made to this data set which makes it really hard to demo with, which is if I was just to um actually

20:43uh let's go over here and let's drag this right up so if i was just to click on this one you see there's actually four rows uh for this order right so

20:51There's there's duplication unnecessarily being introduced here because there's no other information to tell me what the items being returned are.

20:58So all I know is that I've got four rows that say the same thing.

21:02So I think this is a purposely created sort of

21:04a sample data because what we're missing is the information to tell us what's going on.

21:09Either that or we're missing another table that's supposed to tell us what's actually being returned.

21:14that makes sense.

21:14So I'm not going to join that on because it's just going to cause havoc for no reason and it's not really adding value.

21:19So I'm just going to sort of bin that whole section of the flow.

21:22Uh one of the great new things you can do is group things.

21:24So if I right click and uh select those four things, uh this was actually in the previous version you can actually group things

21:31into a nice little folder and you can of course uh name that group so I'm gonna call this connection and join

21:39And now we're pretty much ready to go.

21:40I can of course add a clean step and bring that outside of the uh flow.

21:44What I actually need to do is make sure that uh I bring it um

21:48I have to actually create um a step outside of here.

21:51So I've sort of surlocked myself here a little bit because every time I add something, it's always going to be

21:56inside of that step uh and it's not I can't I can't add anything outside so I've actually just created a container across the whole thing but if you hadn't done that if you just selected the sections that were necessary basically up until here then you would have been

22:09Okay.

22:09So I'll I'll still build this um as as it was, but that's just a a caveat to be aware of with that group option.

22:16So the last thing I'll do is um we'll we'll just quickly add an output here and I'll just go back and check if

22:21this has finished.

22:23Um let's just make sure this is finished.

22:25Yes, it's finished.

22:26It queued for one and a half minutes and it ran in 0.

22:301 of a minute, which I'd assume is like 10%.

22:32So run in six seconds if that makes sense.

22:35So that's that's pretty good.

22:36If I go back to home and I go to explore and then default, we can actually look at my 2020

22:43uh for demo draft here if I click on that we get this interface which shows me the last previous state of this particular demo so if I go back and I hit uh publish

22:54It's actually going to save these changes.

22:56I'm going to get a nice green indicator, and then if I go back and I actually hit refresh this page, you'll see that the draft status disappears from the top and I get a nice clean

23:07uh flow that it's updated to look exactly like how it was.

23:10I have some information here at the top uh about the run history so I've run this once.

23:15It succeeded, it generated 9994 rows.

23:19Earlier on I mentioned the lineage capability.

23:21So here I'm also able to

23:23see each of those uh columns that I've actually got in my data set on the right hand side I've got a little bit more information about what it's connecting to essentially an Excel file the tables in that excel file that it's connecting to

23:34connecting to the actual flow itself, some information about what's going on.

23:38Um the individual fields don't have much information at the moment, but I'm sure Tableau will add to the

23:42this in the future that flow is creating an output here is a data source which is called 2020.

23:474 demo output and it's owned by myself and here if I was actually to click on this I'd see everything else that I

23:53own on Tableau server so that's what the lineage feature does and prep is fully integrated into that so you can really start to see that if you have a full-on workflow in the browser then this is going to actually be really really good for you.

24:04The other thing is it highlights the connection so it's obviously

24:07got my tableau server connection because I output to that you can see that it says that there and my input is actually in Excel over here on the left hand.

24:15hand side.

24:15Okay, so that's pretty much uh prepping the browser in a nutshell.

24:19I'm gonna look over to my question to see if there's everything uh else that anyone has brought up.

24:24Um everyone can hear me, that's great.

24:26Um sessions are C

24:27useful thank you Dimitri um if you've got any other questions about prep in the browser uh now's a good time because the next feature I'm going to go to is spatial features in prep which I'm also going to show you in the browser before we move over to uh web authoring okay so um

24:43Um just looking at Omar, not completely integrated if you use a data source with a contact.

24:48Omar, I'm not sure what you mean by a data source with a contact.

24:54What do you mean a data source with a contacts?

24:56If you can provide a little bit more context while I go to the next feature, then I'll um I'll come back to your question.

25:01in a second okay so I'm gonna go back to uh my view here um I'm actually going to try and see if I can uh if I can if I can uh

25:10uh ungroup everything exactly and what I was supposed to do before is just select these right click and then group and then it does that which is a much better way of sort of

25:19grouping things okay now the next thing I'd like to do is show you that you can actually add spatial files to Tableau prep and before you couldn't really bring in spatial files and join them to your data

25:31And so now you can do that.

25:32To do that, you just hit the plus icon and there's a new spatial file option here.

25:36It's no different to what you'd expect uh pretty much from you know the spatial features.

25:40We've had this in desktop for some time.

25:42Um what I've got here is actually I've got uh two files.

25:45I've got um uh region and state files.

25:47So what I'm going to do is I'm gonna bring in the zip file.

25:50The caveat here is to be bear in mind that when you typically work with spatial files, you typically unzip them and then you connect to the shape file itself.

25:58With Tableau Prep, it doesn't want that.

26:00It wants the whole zip package.

26:01And if you think about it, it makes sense because you're uploading it to the online interface, right?

26:06So you just want the zip file itself.

26:08I'm gonna bring in the state file here and I'm gonna choose for

26:11upload okay that's gonna upload in 3.

26:142 megabytes time so we'll give that a few seconds to just um complete the sort of the natural course of

26:20Action and when that's ready, we should be able to see that here on the left hand side.

26:25So yep, you can see that it's even brought that in for us, and um it's a nice sort of nice touch because it's seen that there's only one table in there

26:32And if I've connected to it, then it might as well bring in the file.

26:35So that's a sort of small quality of life improvement there.

26:38And you can notice that in this preview here at the very bottom you've got the geometry.

26:41So this geometry field

26:43is what we're interested in because I want to join that onto my superstore data file.

26:47Okay.

26:48So I'll go ahead and hit um uh the plus on the clean step and what I want to do is I just want to see which field I can use to join with superstore

26:57Because if I look at the clean step up here in the clean three, I really should be naming these clean steps, but because it's a live stream, I'm sort of pressed for time, so

27:05I'm working slightly faster than I typically would.

27:08I've got a column here that says state.

27:10So ideally I want something like the state name in my spatial file.

27:13And I think I do have that here.

27:15So I do have that.

27:16I have 56 state names there and over here if I go back into this I always like to do the check of the data before and I have 49.

27:24So there's obviously four you know there's a few states that are not in the tableau datasets

27:28that are in the official um geospatial files for the United States.

27:33So we'll be interested to know what those are.

27:34We'll find out in a second.

27:36So what I can just do is I can um just um hopping around the screen here, I can just go in here

27:41and select join.

27:43Now I've chosen to do it there because that's the first available clean step that makes the most sense.

27:47And what I need to do is make sure my join conditions corrected

27:50Because you can see it's not matched it uh correctly.

27:52It's sort of looked for two things that sound the same, but that's not actually what I need.

27:56So again, I just need to join name to state in superstore.

28:00So

28:00So if I go over here on the left, I just choose name, then state and name should all line up.

28:06And then what I'll do is I'll actually get sort of a numbering here.

28:10So I can see here that

28:12Yeah, we we got most of our uh data sort of matching nicely.

28:15Bear in mind that I'm only matching the things in the middle, so things that are in both files, so states that are in both files.

28:20both files but I end up with seven if I just click on those seven you can actually see that there's a few uh states that are in my spatial file that are not in the tableau file

28:29So by clicking on this seven, I'm actually able to see that data here.

28:33And so this is useful context because it lets me know maybe if I'm missing cells from somewhere that we should really have sales.

28:39But it's also just a nice way to get sort of to that, you know, deep data very, very quickly.

28:44quickly.

28:44Uh what I will do is I'll do a left join because I need uh all my data um regardless of if if there's a state or not essentially I don't want to miss out anything that doesn't

28:53have state you'll see my number doesn't change 9994 is still the main number and so we are pretty much good to go what I'll do is I'll move this over here

29:03and I will remove that connection and then I'll just add this to that output so I don't have to go and sort of restructure my outputs uh too much although I actually hadn't uh written anything to this one.

29:13So let's just go uh default and just go um 2024 um spatial output.

29:22And I'll hit run flow.

29:25Uh publish changes, of course.

29:27That always gets me.

29:28I think this should be like, you know, publish changes and run flow as an option if that makes sense.

29:32So if anyone from Tableau is watching, that's uh hopefully some feedback you can incorporate in the future.

29:37And of course it goes back to our queue.

29:39I'm not gonna wait for this to run this time if I just refresh this a little bit

29:43Again.

29:43Did I actually hit run?

29:45It doesn't look like I hit run at all.

29:46I think it published the changes, but did I run the flow?

29:50There we go.

29:51That's been added to the queue now.

29:52And now if I refresh this

29:55We should be good.

29:56Everything's there.

29:57So that's pretty much it.

29:58Those are two new features in Tableau Prep.

30:00That took 20 minutes, a little bit longer than I'd like to, but I just really wanted to show that the whole Tableau Prep pipeline

30:07does now happily work in the browser.

30:09The other thing to bear in mind is that um you know I think there's a lot of people who haven't used WebEdit in a while and so they actually aren't familiar with the amount of connectors that are available through the web now.

30:19So if I just go to this sort of plus icon

30:22All of these connectors you see here, especially these database connections, work in the browser.

30:27So you don't need to do anything like install drivers or get any

30:30anything weird going on.

30:32As long as you have your connection details, you can essentially connect to these as well.

30:36Which is a great real real start.

30:37And of course your static files are still available.

30:40The only downside here is that those static files, you know, maybe you had them synced to a shared point folder and those SharePoint folders were updated maybe by a script somewhere.

30:50You lose that sort of connection.

30:52That's where Tableau Bridge really should be able to sort of step in and bridge this feature.

30:55But at the moment it's not really sort of also improved in its capability and it's not really a viable option.

31:01We don't see businesses using it

31:02a lot which is a shame.

31:03It is a really powerful capability, but it's just not it's just not getting the traction that it deserves for getting static files up to Tableau Online.

31:10So that's something to be aware of.

31:12Databases are obviously much much more

31:14preferred.

31:15Let me look back at the comments and see if uh Omar's maybe got back to us.

31:18Have you tried to create a table with a geometry data type?

31:21How did you set up the um the S rid of the column?

31:25Okay, so

31:28Uh I don't I sorry sorry Omar I don't really understand this question.

31:32Um if I go back a little bit in the lineage you can set up the contact person for an external asset, but if

31:39if you use said external asset in the flow it wouldn't identify the contact person that is true that is true so omar if I just step back through um uh Omar's uh

31:49sort of points here.

31:50If I just go back to uh default and I go to 2020.

31:544 demo and then we uh go to lineage it's a little bit of a Easter egg hunt

32:00And I go back to owners.

32:02The only time you can really know that something's going on is in this page, for example.

32:07What you can do though is you can actually set up data quality warnings and data quality warnings do appear for everyone using that.

32:13data source pretty much across the product.

32:16So whilst um the prep flow is not going to tell you who created the data asset, if I was to go just if I just to just say for example

32:26Let's go to explore, uh go to default.

32:28We created a data source and the spatial output, this one.

32:31I will click on it and I will click on these three dots and I'll add a quality one.

32:36Okay, and I'll enable the warning to say that this has been deprecated.

32:40I'll make it high viz and I'll say live on stream just so you can see this populated and I'll hit save.

32:46That will save the data quality warning to this data set.

32:50Now, the key thing, and I'm not actually sure what will happen here, is if I then go to

32:55a flow now weirdly because I'm on this page I obviously saw that there's a data quality warning there already okay and but I don't see any data quality warning here yet so you don't see it on the dashboard

33:07And I don't see it anywhere here as well.

33:09So on the left hand side it would be nice to have something there.

33:12On desktop you do actually see a little indicator to say there's an issue with this data source.

33:16So hopefully that makes its way to Tableau Prep very, very soon

33:20Um show any mismatch values, yeah.

33:25Uh S-rid the geospatial projection, but for prep instead of data quality warning.

33:29because it's better to use the data roles to cleanse the data right um yeah it is it is a little bit um it's it's it really depends how you how how you sort of do your workflow so if i was to um

33:41This is probably the last question I'll answer on prep before I move on.

33:43If if I just sort of go here and I go to clean three, uh one of the things you can do in prep is assign data roles to specific columns.

33:51So

33:51Think of that as metadata.

33:53So for example, if I just click on on um this one here, uh bear with me one second.

33:59Um

34:00Uh where are we?

34:01Here we go.

34:01So if I click on the icon, sorry, I thought it was in these three dots, but I forget it's actually in the icon where you change data types.

34:07You have this option to assign data roles here.

34:10Yeah, okay, so you can assign specific things that Tableau can be aware that look this is a you know spatial field in terms of geography, email, URLs

34:18Um but you know, this I don't think this is what you're quite talking about.

34:21I think normally in organizations metadata is handled at a much more granular level and you can probably have a lot richer support.

34:27of information in databases that support metadata um sort of alignment and so on and so forth.

34:32So it's again, it's not going to be perfect.

34:35Tableau prep is only

34:37to three years old now and so it's still a new feature relatively speaking new tool so hopefully they add features they are adding them very very quickly and it's something to just look out for and as ask um the ideas page for tableau is very very

34:49well viewed and uh Tableau developers do spend their time on there trying to digest what we're asking for.

34:55Okay, so that's pretty much it.

34:56I've created um a whole uh you know flow here in prep in the browser

35:01I don't want the stream to become just about tablet prep, so I do want to move on to authoring because that is the next big hitter in terms of features.

35:08So if you've got any other questions, please ask them and then I'll come back to them at the end once we've covered everything

35:13just to make sure we get through uh the session today okay so i'll go back into the uh 2020.

35:184 page here i'll actually close this one because i don't want this

35:21uh page and we'll just go back here and we'll go I'm just gonna work my way through the list.

35:26It's actually the easiest way because it's sort of in order of importance as it were.

35:30So the next one is multiple uh mark layer support.

35:33So this is essentially the capability in Tableau to uh do multiple layers inside of um maps essentially

35:41So I can probably hear some of you saying, oh well I could already do do dual axis.

35:45Well this lets you go beyond uh two two layers, okay?

35:48So let's go back.

35:49I'm gonna close this prep flow.

35:50I'm not gonna even save it because

35:52that's all fine.

35:53Let me close a few things.

35:54I'll go back to home and what I'm actually going to do, I'm going to connect to the spatial file I just created in Tableau Prep just to keep this whole theme in the browser.

36:02This is the spatial output that I can uh created.

36:04If I just click on that, I can check that the spatial files actually came through correctly.

36:09I have a geometry field here, yeah, which would have come from the uh

36:13um US sort of file um that I downloaded from the um you know US um geospatial body I forget their name but the geometry field is there and that's what I need so I'll go ahead and create a new workbook

36:26Um this time instead of a prep fly, I'm creating a new workbook.

36:29And you see that it connects to it, it looks pretty much like a an extract published to Tableau Server because of course it is.

36:36Now, the thing you'll miss here is that obviously I've got some um you know data types ready predefined here.

36:42So things like city, uh strings.

36:45obviously measures those are all predefined but everything else is pretty much vanilla I don't have any of the hierarchy I uh metadata that I had uh you know in Tableau desktop for this kind of same file so if you want to create that you'll have to do that

36:57But in order to show you this feature, I'm just going to start very, very simple.

37:00The first thing I'm going to do is I'm just going to try and find the um uh the

37:06location information so there's postal code but there's state here so if I just double click on state this time I'm using the geocoded state so

37:13If you look at this, there's two state files.

37:16There's a state file which is essentially the geospatial geometry.

37:19So you can see that that's here, it's called a geometry.

37:22And this one is actually an inferred spatial field because it understands the list of states

37:26and it's basically uh mapping that to the inbuilt Tableau geospatial um data uh base inside of the product which is actually running on a hyper data source already in the product of that mix.

37:37sense so if I just um expand that you can see here are all the states and they're nicely highlighted what I'll do is I'll make this a filled map just to make it nice

37:45And easy to see, and um I typically add a little bit of uh I call it pop, but essentially I just add the border, a white border.

37:53I meant to turn the uh background uh paste

37:56down not the actual map so let's bring this back up to 100% let's go to map at the very top and let's go to layers and then let's wash out the map almost 80% there we go so we we we can see a bit of the map but I just want to focus on the state

38:10here so you can see the feature uh working nicely the last thing i'll do is i'll uh bring uh state onto uh label um

38:18here okay and so we can just see the individual states that's that's pretty good now what you could have done before and what was always possible is I could have just uh I'm holding command on my Mac it

38:29control on the Windows machine I could have just copied the latitude and longitude and essentially created this weird matrix and then done a dual axis on uh on on each of these

38:41And that would have created the sort of two layer effect that I actually want to see.

38:46But if you want more than two layers, you'd kind of get stuck.

38:49So here you can see now that I've I've

38:51I've I've superimposed both of them, but they both have the map uh capability here.

38:56What I'll need to do is change this one, uh the top one, to a circle.

39:00And what we should see, there is actually a circle there, but um it's um the colour is a little bit off.

39:06So let me uh make this yellow just so we can see it.

39:08um more clearly.

39:09I have a funny feeling it's actually behind.

39:12Yeah that's correct.

39:12It's actually behind.

39:13So what I've done is I've put this the wrong way around.

39:15Classic.

39:16So I'm just gonna try this.

39:18And fingers crossed, yeah, there we go.

39:20So all I did is I switched the pills around and now you can see that I have two layers on my map, but I can't do any more than that.

39:27It's sort of limited.

39:28So I'll create a new sheet and I'll show you the new feature.

39:31I'm gonna hit the state again, double hit state.

39:35So we're gonna get a nice field map.

39:38This time I'm just using the Tableau Geospatial ID

39:41Here, let's go to border, let's set it to white.

39:44And um, I'm not gonna bother with the state names, you can still see a bit of that in the background.

39:47Then what I'm going to do is I'm going to grab this geometry.

39:50Now this has my polygons that I joined inside a Tableau Prep to this data source.

39:55And if I start dragging it, you'll see that you get this new option here which says add a marks layer.

40:01And this is where you need to drop this in order to enable the layer.

40:04If you drop it anywhere else, it will just get interpreted as a normal instruction into in its internal

40:09of a tableau view but if you specifically drop into this gray target area you can see that it's highlighting here then it will actually become a map layer and this is different because unlike before where I could only add two layers I can add as many of these as

40:22as I want.

40:23So let me format these so that's actually clear and easy to see.

40:26Let me grab this one here and I'll just turn it into a circle.

40:30And what actually happens here is that these ones observe the order.

40:34Okay.

40:34So if I change this to orange like I did before, you should see, it's that changed?

40:41Yeah, is it changing or has it changed?

40:43Uh no it's changed, so uh interesting.

40:46I only have one dot.

40:47I wonder why that is.

40:49Um why do I only have one dot?

40:51Why do I only have one dot?

40:53I think that's the aggregate position of all the different um that's the aggregate position of all the different um states.

41:01Let me let me just double check something.

41:03Let me just clear this.

41:04I'm panicking here 'cause I think something's

41:06gone wrong here.

41:07Um let me just double click geometry just to make sure this this geometry information works correctly and then I'm going to drag stay onto layers.

41:16Okay, that works fine and all the dots are there.

41:18And then if I go ahead and add this to orange, that works fine.

41:23So that's

41:23That's bizarre.

41:24Maybe maybe I missed a step when I was creating something.

41:27But essentially if you drag all the items on, then it it just gener generally works.

41:31So here I've not done something exciting because I've dragged two of the same thing.

41:35I'm basically superimposing state dots on.

41:38to state polygon so let's get rid of one of these if I just click on this drop down you'll see that you get a list of options and one of the options is to remove it but I can also move it down disable the selection I'll come to this in a second

41:49second I'm gonna go ahead and remove this initially okay so we just go back to the normal thing I'm actually gonna do something different I'm gonna superimpose the city onto the state okay so let's go ahead and drag city

42:00onto mark layers and what you'll see is now these are meaningfully different bits of information superimposed on each other.

42:07So if I go to color and I just set this to orange so you can get a really good contrast and you can now see this working a lot

42:13lot better and what I can do with this particular layer is I can say uh size this by the sales if I just drag that there and if I just wait a second that will update this is a little bit small so let me just bump this up um

42:29And what will happen after this is that what I want to do is actually add not too much more information, but a bit more information.

42:37Okay.

42:38So instead what I'm going to also do is if I go back to the geometry.

42:42I'm gonna add region onto the color so that I can see the different regions in the background if that makes sense.

42:49So let's just put that onto color.

42:51And we should be good.

42:53So there we go.

42:54I've sort of created something mildly disgusting, and there's a bit of an overlap in terms of colours.

42:59You can see it's a bit of a trip as well.

43:01As I hover over everything, it just keeps flashing.

43:04So what I'm going to do is I'm going to disable this background layer for selection.

43:09Right.

43:09So

43:11What I can't do now is select it.

43:13When you've got your mouse on it, you have to move it off the view and then move it back on for it to work.

43:18Okay.

43:19So if I just move back to uh just reset this again and if I just go this way

43:24Uh come on, let's get this going.

43:27There we go.

43:28If I just move out back this way, you'll see that now as I hover over these ones, only when I hover over the circles does the uh tooltip and everything activate.

43:36Everything else is

43:37just the context so you get some additional functionality that you didn't really have before uh in terms of disabling the layers and you can of course keep sort of keep going um with even more granular information

43:48I think postal code here, if I just very quickly um uh look at this, if I if I just open a new sheet, I'm not sure what this actually is.

43:57actually is in the American dataset is it yeah okay it's pretty good I think we're gonna use that let's just see how granular that is on a map it seems to be wanting to to visualize it so let's have a go and let's let's just bring that in

44:10and drop it into postal code.

44:13Okay.

44:13Okay, yeah, that gets a lot busier.

44:15And it's a lot more busier than the states.

44:17But it doesn't really make sense and in in this particular case because now I've got three things and it's it's a little bit confusing.

44:23So let me just change the shape.

44:25so you can see that they are actually very different.

44:28I'll change it to square and I'll change the size right down so you can see that and then I'll zoom in to a particular section.

44:35So you can sort of see the difference between the cities

44:38and the different postal codes that sort of go with that.

44:41So um you're starting to add lots of information on um this isn't a great example I wanted to keep it simple for the demo just with something that you're familiar with

44:50With so you can try this yourself, um, but I'm very limited in terms of what Superstore actually offers you in terms of visualizing things.

44:57But where this is useful is if let's say you got some spatial information that told you about building locations

45:03Then you got some additional spatial information that told you about the location of roads or rivers.

45:08Then you got another spatial layer that told you the information about a particular area, for example.

45:14look at some context, maybe flood zones or maybe you know fire or safety numbers in the particular area.

45:21You could visualize this context

45:22in the background and then on top of that visualize um the key piece of information that you want.

45:28So this is a great feature because it makes uh layers in maps um data driven which is really really powerful.

45:34It allows you to sort of build that in the same way you know how.

45:37The other thing is that of course you can hack this.

45:40And now a lot a lot of people on Tableau Public have actually used this to build multiple sort of alternatives of charts.

45:47Because essentially what you can do is you can just make any measure

45:50a latitude and longitude and then use that axis um to essentially visualize whatever you want.

45:55So it's it's you're essentially hacking the measure as it were to behave like a latitude and longitude because after all that's all maps are and then you're able to very basically

46:05sort of you know create multiple layers and visualizations to do things you couldn't do before.

46:10I'm not a fan of that kind of stuff.

46:11To me it's a hack.

46:13I know for a fact that Tableau will be adding uh layer supports uh to other mark types.

46:18They've said it uh publicly and that conference

46:19conference as well.

46:20So this is just the first iteration of layers and it's coming to maps first because you know that makes a lot of sense.

46:27But you know in the future of course when they add this to other

46:30capabilities I'd expect them to show some unique aspects to those mark types that you'd want.

46:35So for example, we're dealing with maps here so it makes a lot of sense to be able to disable a selection.

46:40But you can envisage a world where maybe you're dealing with an another mark type and you want to maybe disable something else, like you want to uh change um the way things lay or you want to change the way things sort of augment on each other.

46:51Or maybe you hand you want things to sort of um affect each other.

46:55So if you layer crime rates with I don't know something like um police reports you can have the two data layers interact in a way that's really really

47:04really powerful either driven by parameters, actions, or some of the new capabilities.

47:08So that's that's the flavor of what's sort of possible with this.

47:10It's um to me it's a quite a nice feature.

47:13Most people I think are using it not for maps at the moment.

47:15moment but it does really enable some really nice um options in terms of how you use maps and I think it makes maps a lot more fun.

47:22We're gonna see a lot more sort of creativity in Tableau Public and also in businesses and how they use this and how they implement it

47:29it going forward okay let me check over the comments um omar you're really um you know heavy on the on the comments today it's it's great to have lots of engagement um postal code gives you more than three thousand marks um

47:42Y yeah, um it should do.

47:44It should do.

47:44Actually it's interesting.

47:45I'm on the postal code layer here.

47:47It says eight hundred and twelve marks.

47:49I'm not actually sure what that number refers to though.

47:51Um the easiest way to check is just to go to a new sheet and uh

47:56If you drag anything onto detail, okay, and it will visualize it for you.

48:00At the very bottom it says 630 marks.

48:03So in this particular dataset, uh don't have 3,000 marks.

48:06Um it's just uh you know

48:09uh six hundred and thirty.

48:10The way to break that down would be to add order ID and uh product name, actually product name and order ID because each order can have multiple products.

48:19So

48:19When I add those things to the level of detail, you'll see that this uh number of marks explodes to 9975.

48:26Okay?

48:26So that's that's how to that's how to f uh find out sort of what the level of detail is in your data set.

48:32You know, make sure you understand your data set.

48:33set very very well um uh dimitri high dma has a cost of several hundred dollars plus additional oh data man management add-on okay cool per license um

48:47Yeah, so yeah, this is um this is to do with the data management add-on.

48:51There is actually a per person price that's applied to the whole entire Tableau service.

48:56So if you have a hundred people on Tableau server or Tableau Online and you want the data management add-on, they multiply the price, which is I think uh five or

49:06I don't actually remember.

49:08You can go to Tableau Pricing.

49:09Um, let's go to this Tableau Pricing.

49:14And it says it here in very, very clear terms for teams and organizations.

49:20If I go to this page, so it's $5.

49:2350 per person.

49:25uh per month build annually.

49:26So it depends on the scale of implementation of your Tableau online setup.

49:31So this is all here.

49:32I think it's really easily laid out.

49:35You can sort of get caught out by the

49:37per user per month pricing, but I think it's laid out in a way that allows you to compare it more easily with other products because that's typically how it's you know it's sold in other products

49:45As it were, but Tableau of traditionally always done yearly pricing, but um a lot of competitors do monthly pricing, so that's why I think they changed this format.

49:52You can also see the tableau online option.

49:54Tableau online essentially just costs more because uh your

49:58Tableau's hosting it for you so that the price difference is essentially um Tableau Creator doesn't change, but if you s look at this forty-two dollars for Explorer, 35 the difference in cost is basically the cost per person to host host an AWS setup and make sure everything

50:13things maintained and working properly okay let me go back to um dollars yeah exactly so omar you got the you got the price right there and that's all covered very nicely

50:24For reference, there is a bit of a lag on stream, so I'm reading your comments and you're seeing my response to them like with a one-minute delay, up to one minute delay.

50:33So just bear that in mind if I'm sort of a bit delayed in terms of

50:36Of responding to uh what you see because I might move on, then see your comment in in retrospect.

50:41Okay, so let's go back to the new features list.

50:44Um

50:44The Redshift Spatial Support, this is a sort of a nice one.

50:48It's basically just a very basic function where in Amazon Redshift, which is a data store type, you can now add spatial information and Tableau can interpret it.

50:58Traditionally, Tableau's always been able to connect to Redshift, but what it didn't pull in are the spatial fields that were actually held in that.

51:05So there's nothing really to show here.

51:07I could have set up an AWS instance just to show this

51:10But I think we all know what spatial files look like.

51:12We've just worked with one.

51:13And the connectors are always very boring to demo because um if they work, they work.

51:18You've connected to your data.

51:19The analysis is always where people

51:21sort of get excited.

51:23I've obviously shown you the spatial support in Tableau Prep here already, so that's really good.

51:27The offline map for Tableau Server, this is a quirky one.

51:30This one is essentially when

51:32Let's say you work in a bank and some banks have this really sort of strict requirement where that they're where their Tableau server is completely air gapped from the world.

51:40What that means is their Tableau server can't communicate with the internet.

51:43It can communicate in their private network, so it can communicate within the bank, it can communicate with other servers in maybe other locations, but it can't communicate with the internet.

51:53And the issue there is that the Tableau mapping assets

51:55the you know the tiles that load up when you load a map um they actually need to load from the tableau server.

52:01So what they've actually enabled is the ability to create offline map support but the

52:06The key thing here is it's actually available in web edit.

52:09So so it's not it's not that it wasn't available before, it's always been available, but it's actually now available in web edit, which is a slightly slightly nuanced sort of feature.

52:18To me, I don't even think this is a feature.

52:20This is a this is a quality of life improvement.

52:22This is an enhancement or a fix really because it should always it should have always been there if that makes sense.

52:27Okay.

52:28The next thing is a resource monitoring tool for Tableau Server on Linux

52:31The Tableau resource monitoring tool has existed for Windows, but I think two to three years ago Tableau switched, um didn't switch, they enabled the ability to

52:40install Tableau Server on Linux.

52:42Now I'll let you into a secret.

52:44Whenever I've installed Tableau Server on uh Linux machines they just run so much smoother.

52:48There's just there's none of this sort of faff you have with Windows servers where there's you know there's this

52:53this plugin or this driver here or there's this service here hogging this resource.

52:57The Tableau servers that run on Linux just tend to run faster and I'm pretty certain Tableau Online itself is probably run on a cluster which is based on Linux

53:06not Windows.

53:06So uh if you're still on the Windows train because the organization's on that something to be aware of.

53:12But this was actually a really important

53:14functionality that uh server admins really like to use to make sure things are you know working okay and um having it on linux now means that that this this parity between the two

53:24um and uh admins have an easy place to go see things.

53:28Now it's not to say that you can't collect this information yourself.

53:30There are tools that can collect this for you and present it in a slightly different way.

53:35A tableau is just offering you the ability to look at this

53:38more easily.

53:39It's part of the server management add-on.

53:41So you can see here that as with most things recently, there's a there's an add-on for everything

53:46So there's part of the server management add-on.

53:48So there's Tableau Server, then there's the server management add-on.

53:51If I go back to my pricing page, you'll see that that's

53:54three dollars per person uh on your uh entire sort of licensing setup so you've got to be aware of that but otherwise it's pretty much um

54:03Exactly the same as what we're used to on Windows.

54:06If you haven't checked out sort of Tableau server features, that's also fine.

54:10It's actually a very easy thing to get into.

54:12There's some really great documentation.

54:13Tableau have some of the best guides on their website to help.

54:17Now the next thing is web authoring enhancements.

54:19Now this is a weird one to cover because there's actually quite a few things in here.

54:22So I'm going to try and work through this very

54:25Um I'm gonna say quickly um but um I wanna I don't want to sort of skip over everything because nothing here is new it's just that it's available in web edit, okay?

54:34And so if I just very quickly uh go back to my map which I was building before, I'm just going to create a new sheet and I'm just going to do a very, very basic chart.

54:42I'm going to do a chart that I know I might need later.

54:45So let me just change this to a continuous month.

54:48And let's change this to um uh sales uh so we have a nice little time chart.

54:54Okay, so if I go back to this, let me actually put it in a new window so that I don't keep swapping the screen and uh

55:00frustrating everyone flashing between tabs.

55:03The first thing is highlight actions.

55:05Okay.

55:05So highlight actions um weren't available through uh web edit.

55:10So the way to do this, um this is obviously the worst type of chart to show though

55:13that actually so I'll keep that open and I'll just bring in uh let's bring in the subcategory onto color and let's just bring uh profit onto rose and uh let's bring uh

55:27Um what should we bring across the top segment?

55:30I think segment makes sense, yeah.

55:32Cool.

55:32So we've just broken down some segments and we're looking at profit.

55:35It's not exactly a beautiful chart, but what I want to do is show you the ability that you can

55:40you can use here which is uh over here on the right hand side to enable a highlight action you just need to right click on the actual pill you show the highlighter and it appears at the very bottom

55:49Uh you can't move these around unfortunately.

55:52You can on a dashboard but you can't in the in the browser, so they appear where they appear if that makes sense.

55:58If you don't want the label you you can't unhide the label, you can in desktop

56:02can't hear so the thing here with the highlight action is you obviously you just uh choose the thing that you want and it's obviously highlighting the chart on the left hand side that's all it really does when you select what you want it keeps it highlighted and your mouse

56:14can sort of move around and free.

56:15That's now available in web authoring.

56:17The next thing is mark labels.

56:20You can quickly make adjustments to this.

56:22This is actually a quirky one.

56:24If I just uh if I just create a tree map, let's just put um

56:28Sales on site and I will put the detail to be let's say category.

56:37Where's cat mm

56:39Yeah, let's put category on detail.

56:40So that should give us a tree map, great, and let's put color label show mark labels.

56:46Um essentially you have this capability here to uh

56:51to basically sort of control them a little bit better.

56:53It's a lot easier to control them and they work a little bit smoother as well.

56:57I think there's there's been some some notes that actually across desktop and um web authoring

57:04tool tips and uh interactions are a little bit crispier as well so that's that's something to be aware of you can create extracts so this is a quirky one to show you I'm gonna go ahead and connect to a new data source I'm gonna hit the plus icon

57:16And obviously here you see our published data sources.

57:19If I go to files, I can upload a file and just for people's reference you can actually connect to quite a large range of data sources.

57:26I don't think a lot of people are aware of this.

57:28A lot of people think Tableau Online is purely uh, you know, just for connections to data sources that have already been published.

57:35That's absolutely not the case.

57:36You can connect directly to databases.

57:38It's not as big as uh desktop, but of course um the feature is there and obviously these these are gonna continue growing.

57:45Over time, let me go ahead and upload a file.

57:47Let's go to desktop, let's go to my live stream folder.

57:50I'm going to choose the same superstore sales files that I was using before.

57:53And as we upload this, um

58:01Uh let's just wait for that to be done.

58:03As we upload this, we get here, okay?

58:04And you probably wonder, well, uh, how does this work?

58:07So let's bring orders onto um

58:10I pulled into the wrong place.

58:11I think I think I think I thought I was in prep.

58:13Let's pour this onto this canvas and I'm just going to click update so we can see the table that we are used to seeing here below.

58:19And of course you can see this option here on the right hand size that says extract

58:22Now if I click on the extract, you'll see that extract will include all data.

58:26Extract will be created automatically when you switch to a different data source or when you switch tabs.

58:33Okay?

58:33So

58:34The key thing here is that the extract doesn't get created straight away like on a desktop where you hit extract, you choose some filters, boom, and you're done.

58:43Okay, everything actually needs a little bit of time to process because you're on server, you're online.

58:48So if I go to a new tab here, what will actually happen?

58:51is it will actually start creating the extract.

58:54And in this particular case, you can ask it to notify you when it's complete.

58:58Okay.

58:58And when you do that, you it's gonna ask you where to save it.

59:01So let's say extract demo

59:05Into default, set save location.

59:07You can now close the editor, and when the extract finishes, the workbook extract demo will be saved to the default folder.

59:13Okay, so

59:15That really breaks my flow because what it then does is it chucks you out of the experience.

59:20This is consistent.

59:21This is not like a bug.

59:22This is not

59:23something that I may know made a mistake.

59:24This is what it does each time.

59:26So it's a little bit frustrating because I was actually in a workbook.

59:29I was in the flow of doing something.

59:31So I always recommend you do that like once you've actually finished building what you

59:36need to build and um you you can sort of more easily work with it so let's go back into it and I'm actually gonna go back into uh sheet one I think this is the one I created I hope this is in the previous demo yeah this is the one I created earlier on let's go hit edit

59:50Um the extract is being created in the background.

59:53If I go back to sheet six, um you'll see here that the extract has finished, okay?

59:58So it's over here on the left hand side and essentially it was running in the background and it's in that same place I showed you earlier for Tableau Prep.

60:06So if you're just joining the stream now, earlier on I showed you that you

60:09you could go to the jobs list and essentially check all the cued jobs that are available on Tableau server.

60:15You can go back to the beginning of the stream to sort of see that.

60:17But here the extract is finished and now we're pretty much

60:20good to go.

60:21So extracts on the browser, they're a little bit clunky.

60:25They don't quite work like you used to, but they work.

60:28Everything gets added to the queue and once that's done you can obviously come back into your workbook and start finishing

60:33You can be notified by email when it's complete as well, which is a nice quality of life improvement as well.

60:38Apply filters to worksheets.

60:40This is um this is an interesting one.

60:42So if I was to uh let's say build a dashboard

60:45This is the worst dashboard I've ever built, honestly.

60:48Um let's uh don't no one please take a screenshot of this and use this as an example.

60:54Uh colleagues, if you're listening, this is not an example of my uh uh

60:59uh work that I deliver for clients, but we're gonna go with it because um I'm trying to I'm running out of time actually.

61:05Um I'm not I haven't got enough time so I need to speed up if I'm gonna cover everything

61:08everything.

61:09Um the key thing here is obviously if I go back to this sheet and let's say I add a filter here.

61:14I don't have anything in the filters plane.

61:16So let's say what shall I filter with?

61:19And I want to do category.

61:21Let's choose country region.

61:22That's a pretty good filter, yeah.

61:24So let's um let's just add this here.

61:26I'm gonna put in the filters pane.

61:28I'm not gonna bring it into the view just yet.

61:30Uh of course there's only one country.

61:32I actually meant city

61:35Let's click all, click OK.

61:37Then I'll go back to my dashboard.

61:40So you'll see here that this dashboard.

61:42If I go down to this drop-down list and go to list of filters, I can bring in the different filters.

61:47And if I bring in City, you'll see that I get the long list of items.

61:51Now the thing they've added in uh 2020.

61:544 is this full filters list.

61:57This wasn't actually available before.

61:59You had a limited set of capabilities.

62:01You could edit the filter.

62:02Um, you couldn't really do much else, uh, to be honest.

62:05But now what they've done is they've actually updated this.

62:08So now you can even choose whether to apply this to

62:10all items in the worksheets just stuff on the dashboard selected worksheets if you do that you get this full interface that you're used to seeing before uh in desktop so now this actually makes it a viable option to go into Tableau server or tableau online and fix things very quickly because

62:25You're now able to do that very easily if it's a filter that's sort of causing you uh heartache, if that makes sense.

62:30Okay.

62:31The last thing to do is create fixed

62:34So what I could do here is let me just set this to take up the full screen.

62:41What I want to do is I'm just gonna select this item here and I'm gonna hold command and select um

62:48Uh the knolls are behind my head.

62:50Uh thank you, Lorda.

62:52Uh I uh Lorna, I I can't um I'm because I'm on a lag, um things things are sort of working at a slightly different pace.

63:00So

63:00Um I think I'm seeing your messages but I'm seeing them late if anything so.

63:04Um at some point I'm gonna look back on this live stream and play back Lorna's comments and

63:08I'm gonna be stuffed.

63:10It's gonna highlight everything I got wrong.

63:12Um but anyway, I'm gonna go ahead and create a set.

63:14This is something now you can do here on uh the browser.

63:17So here click create a set.

63:19Now when I create that set, because I've done it here in sheet six.

63:23Um if I go into a new view um and actually go back in here, you can look for something called set.

63:30Now um let me just actually call uh let's see if I can

63:33can finder here so you go category set so it takes the name of the field and uh uses that to basically create the set if that makes

63:41sense and if I right click on that there's no ability to edit things so you can't add items to the set from here in order to do that what you actually have to do

63:49do is right click uh an item so if I click on this item and uh go back down to this option I can add to the existing set again or I can remove this item from the set if it was part of it

64:00So unlike on desktop where you can control the sets in a separate interface over on the left-hand side in the data pane, here you can't do that.

64:08You have to do it basically through the data source.

64:11I keep saying right click, but what I actually mean is normal click.

64:14And then you get this option here to add or remove items to a set.

64:18Okay.

64:18So if I just bring back the uh features just to summarize them very, very quickly.

64:23Highlight actions, mark labels, create extracts, apply filters to worksheets, and create fix.

64:28sets that's everything um uh thing uh lawna's saying that comment was from a long time ago with the maps uh I I did figure there were nulls missing somewhere um but I yeah I

64:39I just I decided to ignore them.

64:40I'm sorry.

64:41Um now the next set of features, actually two features I can demo very quickly, and actually we've we're actually okay in terms of time.

64:48So

64:49We're doing pretty well here.

64:51We've got another half an hour or so to go.

64:53So let's let's look at predictive modeling.

64:55Now I'm not a stats boffin, so I'm sort of fully putting my hands up here and saying anything I say in the next

65:01you know, 10 to 15 minutes while I talk about this and Data Access extensions should not be used against me in the future for the record.

65:09So

65:10Um, Tableau have been slowly adding predictive modeling functions inside of Tableau.

65:15In the previous release, they actually added some capability to uh essentially

65:20essentially use uh predictive modeling functions to create um essentially how do I explain this the the create the added functions but the functions themselves allow you to create sort of new contexts in visualization.

65:35By new context, I don't mean that it's, you know, trying to run a machine learning algorithm to predict X, Y, and Z.

65:41They allow you to add context like trend lines that maybe

65:44you know following a certain predictive model based on some sort of type of predictive analysis like linear regression or something like that.

65:51And so it's actually been really nice to see them slowly add to this.

65:54because it's actually started to become quite a fit sort of well thought through set of features that doesn't take predictive modeling into that sort of really vague area where you don't really know what's going on.

66:05um but it keeps it simple and applicable to analytics and charts which is um what were what I'm gonna show you.

66:10The other thing is whenever you're doing predictive modeling into the future as it were and you need to see you know a month or two in the future

66:17You always had this problem in Tableau where the date access was limited.

66:21You can actually go forward in time and let the model sort of run through in time.

66:25So what we have now is this new feature as well in 2020.

66:294, which allows you to add a date access.

66:31extension.

66:32So I'm going to show you both of these features in the same demo.

66:35I'm going to go back to the browser and we're going to um have a go at that.

66:39So I'm going to actually close this and just start from a new data set.

66:43Don't think

66:44I want to stay in the same workbook because it is super messy.

66:48So let's just go back and uh it goes back to the view of course.

66:51I'll go to the default uh folder.

66:54And we'll uh let's let's let's let's use our 2020.

66:574 demo output from the first moment and let's hit work

67:02open.

67:02Okay, so here we are.

67:04We're in a plain vanilla workbook.

67:05Let me go ahead and add order date to columns and I will go down here and make this

67:11a continuous day actually I need a continuous month uh it doesn't make sense to be doing predictive models on a day to day basis uh months tend to be uh more

67:20sensible periods of time and then I'll put sales onto the view here.

67:25Now the thing to notice here is that when I go down and

67:29uh click on this um they call it the carrot this little down arrow um is called the carrot sometimes um by tableau uh professionals um you'll see this option here to extend the date range okay and i can choose how many

67:41uh weeks, months to or years to extend this by.

67:45Now I can extend it by six months and you'll see that actually nothing happens.

67:49That's because there's nothing actually really requiring this to sort of

67:52Go forward because that only really works.

67:55The data access extension and the predictive model really only really start to kick in when you're using table calculations.

68:01Okay, because this sum of cells is not a table calculation

68:05it's not going to do anything.

68:06So a quick way to change this to a table calculation to show you what I mean is to do a moving average.

68:11So I think I set this to let's just see how long I set this to I need to remember.

68:15So six months it's

68:17looking at if I change this to a moving average you'll see that my dates go out to March 2021.

68:24Now it's a very subtle change.

68:25Let me do that again and go back.

68:27If I go back, my data set ends on December the 20th, uh December 2020.

68:32If I go forward, my moving average uh stops in February 2021, but my date axis goes all the way to

68:39I think it does actually go a little bit further to March, okay?

68:43And the way to sort of check this, if I go to edit table calculations, you can see that it actually looks at the previous two values

68:50And so the reason this stops in February is because it's only looking at the previous two values, I ran out of values uh in two months.

68:57So if I was actually able to set this up to six.

69:00Um what it would do is it would extend the date range up to the sixth month, if that makes sense.

69:05If I change this to seven, okay

69:08It does keep going, okay, but what starts to happen is I start to lose uh context out of my data set, if that makes sense, okay?

69:15So um it doesn't really make sense to sort of misalign those two values

69:19If you're looking uh forward, um if you're looking if you're doing a moving average of roughly six month app moving average, you need to make sure that your day access range looks forward long enough

69:30for that to not look crazy but also show you enough um context for the future if that's if that's the way you want to run this okay so that's that's a data access extension it's very simple it only works on table calculations

69:43Actually, let me rephrase it.

69:44It only works on certain types of calculations.

69:46If I actually go back to the um, where is my uh what's near in 2020.

69:514 page?

69:51So uh date action date date extension works

69:55with predictive modeling functions, table calculations like running average and external services like R and Python.

70:01R and Python makes sense because of course you could be doing predictive modeling in R and Python.

70:05And so like Tableau's own predictive models it makes sense

70:08sense so table calculations uh predictive models r and python are the sort of two instances where they work so i've shown you the data access extension

70:17Let me show you predictive model and let me show you how those two work in synergy.

70:21Okay, so what I'm going to do, I'm going to clear this table calculation.

70:24I'm going to remove this and I'm also just going to move remove this data as this extent.

70:29So that I've got a nice reset point and you can see me do everything from scratch.

70:33I'm gonna add the sales in here.

70:34We're just gonna keep this uh very simple.

70:37I'll go down here and I'll create a calculated field.

70:39Okay, and what I want to do is actually bring in some of the predictive modeling functions.

70:43Now I'm a real hack here.

70:44I don't I don't remember functions at all.

70:47I've become 100% reliant on the fact that this little sidebar is always there.

70:51If that surprised you, if you've never seen this, I I keep meeting people who've never seen this little arrow here on the right hand side or thought to press it and found that everything is actually really well described here on the right hand side.

71:01side.

71:01So let's um let's just type in model and you'll see that there's actually two functions.

71:06One called model percentile and model quantile.

71:09Now this is where my predictive

71:10and statistics knowledge is gonna really sort of um you know put me in a corner because I don't really know how to explain these well but I always give it a go.

71:18So

71:20Model quantile.

71:22The best way to think of model quantile is if I take something like the median, the median value is actually part of a quantile, so it's the middle value.

71:280.

71:2925 is the lower quartile, as 0.

71:3275 is the upper quartile.

71:33So quartiles are like ranges.

71:35They essentially represent roughly where your data might be in a specific range.

71:41So if I take the median value, that means I'm tracking essentially the middle of my data set

71:46Okay, so it's working on within a certain range.

71:49The percentile is slightly different because what it does is it gives you the probability of a certain percentage of data being above or below the line.

71:59Okay, so those two.

72:01They're sort of they are different functions uh but they have slightly different purposes in terms of how they work inside of Tableau.

72:08So that's that's the simplest way I can come up with describing it so far.

72:11I'm still trying to improve on it.

72:12So if you've got a better description, you know where the comments are

72:15let me know but today I'm gonna use a model quantile and I'm just gonna actually double click this so that it adds it nicely for me so I don't have to type it

72:24Now when I click inside of the function, let's see if I can just expand this.

72:27I think it's going to expand everything.

72:29Yeah.

72:29Oh, I'll expand everything actually.

72:30It's fine to be one size bigger.

72:33If I

72:34Type this in, you'll see that it gives you an example of how the function should be written.

72:38Okay.

72:39And if I just um if I just sort of uh

72:43Look at this for one second, you'll see that I don't really I don't I have two options.

72:48Okay.

72:48I have this one here which is model quantile target expression predictor sessions.

72:53Or there's a little there's a little tiny or here that you can't really see says I can either describe it like this where I tell what the model is, the quantile target, and the expression predictor expression.

73:03Okay, that's because there's two ways of writing the predictive models.

73:06The first way is to just assume you're using the default predictive model, which is the linear regression, and that just means you don't need to specify the model that's being used

73:15whatsoever then I use a target expression and the predictor expression so let's go ahead and write that first I'm gonna find the median of my data set so I'm just gonna type in uh 0.

73:245 here so I'm just gonna type 0.

73:265

73:27I'm looking at the reference to see if that needs to be in speech marks.

73:29I don't think it does.

73:30I think I get away with it.

73:32The predictor expression is essentially what I'm trying to predict.

73:34So in this case is the sum of cells.

73:36So let's just type uh some, let's hit enter of sales, let autocomplete do its work.

73:42And then the last thing I need is um um uh sorry, I think I said predictor expression previously.

73:49What I meant is target expression.

73:50So the thing I'm trying to predict.

73:52And then the predictor expression is basically this here at the top, the date, because that's actually what's going to control how the model works if that makes it.

74:00sense so if I just type um I could type it but there's a there's gonna be a problem so if I just type in order date um there's a there's a there's a real big problem with this which is that the order date is at a day level okay

74:12And secondly, I also have this issue where if I just click on this uh error here, I can't mix aggregate and non-aggregate arguments with this function.

74:21Okay, so I've got two problems to solve here.

74:23I'm gonna solve them uh in a very cheeky way.

74:25I'm actually gonna delete order date there and I'm gonna hold command and drag month

74:30uh order date here because in my calculation window what I actually have is not order date what I actually have I'm trying to expand this but I can't um what I actually have is um

74:41month of order date.

74:43So I need actually a date truncation function to get this to work at the same level of detail as my view.

74:48And then to get rid of this error here with aggregate and non-aggregate functions, I'm just going to make this into an attribute

74:54So ATTR, go over to the end and just get this going and hit apply

75:02And now that works.

75:03You can see that it's appeared here on the left hand side as calculation one.

75:07I'm gonna call this um median uh predictive.

75:12model.

75:12That's not what it is.

75:13I should call it quantile.

75:15Let's just stick strictly to the terms that I know.

75:18Hit apply.

75:20And then hit OK.

75:22And what we're going to do is we're going to uh drag this out.

75:25And I'm not going to drop it here.

75:26What I could do is put it here, and then what you see is that it creates this line and they have you know generally the same axis.

75:32What I'm actually going to do, I'll show you a little trick here.

75:34It's a nice quick way of getting things into uh into the view if you want them to sort of be superimposed on each other.

75:40Um of course it's not going to work for me because I'm trying to do a demo.

75:44It probably doesn't work in Web Edit.

75:46That's where I went wrong here.

75:47So you can see I'm trying to drag this here into the axis.

75:51It's not letting me for some reason.

75:53It can let me drag it into view, but it's not letting me drag it here.

75:55So I will have to put it next to cells and do a dual axis and and what I'll then do is synchronize these so you can actually see the extent.

76:06I keep right clicking and um

76:09Oh no, I think I have to synchronize it up here.

76:11There we go.

76:12Dual access.

76:14Uh no, okay, the synchronization option is not here.

76:17Why is that missing?

76:19Um why can I not synchronize this

76:23What am I missing?

76:26No.

76:27Let's just check this again.

76:31Am I being blind?

76:33No, I don't think I am.

76:34Dual access, I should have a synchronization option here, but I don't seem to have one.

76:39Um I can't explain that.

76:41Um I could swear that existed before.

76:43What I will of course do is open up

76:452020.

76:464 um desktop and basically rebuild the same thing while that's opening let me edit this and grab the calculation um

76:59That is probably another web edit limitation, um Steph, if I'm honest.

77:03Um But I swear I've gotten past that before.

77:06I don't know what's going on.

77:08I think it's just uh you know demo bugs as it were.

77:11So let's go ahead and connect to super

77:13or US.

77:14I'm gonna go ahead and create a very quick calculation here.

77:17Just uh the calculation window came off screen.

77:19This is the exact same calculation I created before.

77:22So call this model quantile

77:25And now I'm here.

77:26You know, I could have I could have done my little trick anyway.

77:30Oh god, uh demos, demos.

77:32If I do a one and a half hour demo real, something is bound to go wrong, isn't it?

77:36it so I guess I was asking for it.

77:38So let's let's go ahead here and choose month and let's get um sales back in here because this is what we were looking at before and what I can do you can see here that I can actually drop it here onto the access

77:51Whereas before it wasn't letting me do that.

77:52So what I did before is I did this and then I dual accessed those two.

77:57And what I'm looking for is this option to synchronize the access, which

78:00wasn't showing up on I swear that's been there before.

78:03Maybe I'm maybe I'm imagining things that sometimes that happens.

78:07So what I was looking to do is synchronize the axis like this so that you could see that this line was tracking the median

78:12fairly accurately.

78:14Let's remove this model quantile and I'm going to drag it on again, this time doing what I was planning to do before, which is to drag it onto the axis.

78:21And when I do that, you see it does uh something slightly different.

78:25It does multiple measures in the same um in the same.

78:28I never know how to describe it.

78:29This multiple measure, I'm just gonna call it multiple measures in the same um measure values pill.

78:35Okay, so uh it's achieves the same effect, but it's much much faster because it's just one drag and don't have to mess around with

78:42uh dual axis seeing as the axes are showing the same thing that's the caveat you can't really do this when things are showing different things at different scales it has to be really literally the same thing on the axis for both of them for this tool

78:53work really well so there you go you can see this uh very nicely now what I'm going to do very quickly I'm gonna do a little hack here I'm just gonna very quickly edit these other ones to show the 25

79:06Because before what I before I um before I show you the predictive modeling capabilities, which is where I'm sort of heading to, I need to create the lower quarter and um

79:18If I edit this one, uh I need to create uh 75.

79:22Let's create this, let's call this upper quarter.

79:29Click apply and um uh quantile uh let's actually call this what it is, um median

79:40Okay, and what I can now very easily do is I can drag the model uh lower and upper and drag them in here and you'll get a line at either side.

79:48Okay, so now you can see the predictive modeling functions working, but they're not doing much predicting because I've not extended the date axis range.

79:55Okay, so let's go ahead and do that.

79:57Let's go ahead to date and you can see here I've got the option to

80:00extend the date access range and if I extend it by uh six months you'll see that it basically carries on it just extends the line now this model I've got isn't very good

80:10Because if I extend this to a year, it just takes the trajectory that I had over here and just literally sends it off.

80:17But notice as the time period gets bigger, the m the gap between the upper and the middle quartiles change.

80:23And that's because

80:24The predictive model is trying to predict what might happen in the future based on what sort of what happened here.

80:29And it's essentially saying that the further out you go in time, the less accurate that's going to be.

80:33So my data's going to fall inside of a bigger and bigger, bigger range.

80:37makes sense.

80:38The other way to show you this is if I set this to some crazy value, let's say set this to two years, right?

80:44Just to really exaggerate this.

80:46So I've got this option here where you can custom choose a unit of time

80:50and a period of time so I'll just set this to let's say let yeah let's set it to two years okay and what I will do is I will add a month to filters pane okay and I'll do range of dates because

81:03that's what I actually need.

81:04I always forget the trick of right-clicking the date and just showing the filter to make it work faster essentially.

81:11And watch what happens as I sort of reduce this filter, right?

81:14So if I reduce this filter, notice it's still extending the range

81:17Okay, uh by a certain amount of time.

81:19Um but um as I do this the prediction goes wilder and wilder.

81:24So basically when I have such a small extent of data, the orange line is my actual data

81:29These lines get wider and wider and wider to the point where you know it's just the the range is so wide this is this is garbage.

81:36This is not something you can use um in any sort of visualization.

81:40But I I a s I think if you're actually smart with two things.

81:44If you're smart with the uh date extension, so if I just say three months, I think that's actually quite reasonable.

81:49Um if I've got this sort of long history of data and some seasonality to it and I extend it by three months.

81:55That makes sense.

81:56That's the first thing I can do.

81:57Choose a w a reasonable day access range.

82:00The next thing I can do is choose a better model.

82:02And that's where we come to a new feature in 2020.

82:05Yes, I'm still on predictive models.

82:07Um, we've actually got some new predictive modeling functions.

82:10Okay, so previously we only had linear regressions, but

82:14We now also have the ability to add a and I can never say this uh linearized regression

82:23No, that's that's just so wrong.

82:24Let's let's try that again.

82:26Let's go to um uh let's just tableau predictive models.

82:31I swear I can never say this properly.

82:33It's an absolute twang tongue twist

82:35So let's see, choosing uh predict choosing a predictive model.

82:38So you've got the linear regression, which is what I said before.

82:41It's a regularized linear regression.

82:44That's a real tongue twister.

82:45I I'm can't believe I even got it right first time there.

82:48Um but notice that the examples they give you here are actually writing in a slightly different way.

82:52different way.

82:53This was actually the second way I highlighted earlier on.

82:56So this model linear uh model um uh uh RL which is regularized I can't say it now regularized linear uh uh regression I've read that as aggression for a second.

83:07I think I'm getting aggravated now.

83:09Um, and then you've got this last one which is uh the Gaussian uh process regression, which is the GP one.

83:15I'm actually going to choose this one.

83:17I'm going to copy this terminology.

83:18I'm going to add it to the front of my calculation to change these models, okay?

83:23And what I'm going to do is I'm actually going to pause the view just so that you can

83:27see the effect a little bit uh better and uh what I'm going to do is essentially just need to add a string uh to the front of this like that so that says model gp hit apply the chart will go gray because I'm changing the calculation and I

83:40paused the I've paused the um the view so if I hit paste on that one hit apply hit okay and the last one do the same thing

83:51Paste that in, hit apply, hit okay, and then hit play.

83:56You'll notice that this model does a lot better job at noticing seasonality, okay?

84:03And uh it is still sort of extrapolating into the future, but you notice that it's it's doing a much better job of it than uh the previous one because it it's it's able to sort of notice some uh a level of

84:14seasonality remember the upper quartile is this top one here and the lower quartile is this red one and uh the blue one is the uh median median so

84:23of predictive value so this actually starts to work a lot better because um if i just get rid of the if i just get rid of the um upper and lower

84:32Okay, and I show you what a trend line does.

84:34You'll you'll notice that this um okay I can't add uh I can't add a trend line.

84:39Uh here I can.

84:40Uh not from the axis, but I can do it here.

84:42See a trend line just generally draws a line in the general direction of the data set, okay?

84:46And I I I do have some control on the trend line.

84:50But um these predictive models I think do a better job of tracking where the data's likely to go.

84:56So obviously it's noticing there's some seasonality around winter and uh you know the back end of summer here and it's uh sort of

85:03pushing that through into our data set.

85:05So it's obviously looking into even though my value here in June is really low, it is looking into the past and suggesting you're likely to perform better.

85:13So this this model I think is actually starting to show something a little bit more realistic and I've not gone too wild with the data extension.

85:20So it is starting to give us a little bit of context of where the data might go.

85:24So people are familiar with, you know

85:26know what's going on.

85:27What I'd like to be able to do is just to change the color on everything that is part of the predictive model on this line.

85:35It's

85:35It's a little bit quirky.

85:36I don't know how to figure out how to do that.

85:38I think what I'd have to do is somehow figure out a calculation that knows what the maximum extent of the data is, then changes the color of the line.

85:46uh manually but what I'd like to do is have an option so that I could either make this line dotted or just change the color to let people know that this isn't real data this is extrapolated data based

85:56on what we know.

85:56So if you were going to use this as a some sort of guide, the last thing you do is maybe change the color so it's more muted, hit apply, and um there you have your context, you have your predictive model.

86:06working inside of a tableau view.

86:08And I think this is a a very light and appropriate use of predictive model in terms of analytics as well.

86:13So that that was um

86:15That was that was a really long demo.

86:17Gosh, that took 15 minutes to show you those two features.

86:20But they need a lot of detail.

86:22I appreciate predictive models in Tableau aren't something that everyone's covered.

86:25So I really appreciate that.

86:27I haven't looked at my comments in a long time

86:29Um uh Omar, yes, it does work in desktop.

86:32I think there's a buggered browser, uh, to be honest.

86:34Uh that's another advert for how much work Tableau needs to do to bring that up.

86:38feature parity.

86:39Um it seems like the forecast option gives you more capabilities with seasonality.

86:43Yes the forecast option does um but the

86:46The forecast option uses the standard uh forecasting model that's used in R, I think.

86:53Um, and it that's actually quite a limited model.

86:56So the predictive models um

86:59G give you an alternative because number one you have a choice of predictive models.

87:03Number two, it seems like they're using uh they're gonna use this architecture to bring more types of predictive models to those functions.

87:09So they've basically built the you know the piping and they're now gonna add flesh to these

87:13features over time the date access extension was an important part of that um but you know i can't i can't speculate that's sort of just what i'm thinking based on tableau's history of releasing features and how they typically

87:24Work okay.

87:25So um let's go back to my browser.

87:28I've really I've honestly forgotten what else we have to cover here.

87:31So we have that was the tableau pricing.

87:33Where is my tableau new features?

87:35Here we go

87:36So we've covered predictive modeling and data access function.

87:39Custom views tab.

87:40Okay, this is a nice easy one to show you.

87:42This is um a nice change of tone after covering predictive models.

87:46Wow.

87:47If I go back in here to the home view and I just go to uh superstore workbook, there's got to be a workbook in here that I can.

87:56look at.

87:56I'll look at web authoring improvements and what I'll do is let's let's look at this as a grid view over on the right hand side and I will go to dashboard one which is another ugly dashboard that I built a long time ago.

88:10go to demo a feature.

88:11This looks more coherent than my previous one though.

88:14And what I'm gonna do is I'm actually going to change uh this filter here to just consume it.

88:18And it's going to change the charts a little bit.

88:20And what you'll see is that at the top we have this toolbar.

88:24And this toolbar is essentially where you have certain functions that relate to the visualization.

88:29If I go to uh view, uh you'll see that I get this custom views tab, okay?

88:33And I can go ahead and create a new one.

88:35So I'm gonna call this um consumer view.

88:39And what this does is actually remembers this uh setting for me.

88:42If I choose this as make it my default and make it visible to others means others can see that I've created this

88:47consumer view and they can use it too.

88:49So when I hit save notice two things the viewer refresh number one number two the default view that I'm looking at now is view consumer.

88:57So if I was to go back to default

88:59and then go back into web ordering improvements and go to this dashboard.

89:03What it will take me to is my consumer view by default because I selected that option at

89:08before okay now the thing is users create this all the time but admins want to know what people are creating so they can see if there's any insights or views that people might like to see more often.

89:18So

89:19There's a new custom views tab where I can actually see all of this in one place.

89:22That's pretty much the feature in a nutshell.

89:24You can see I created this uh 30 seconds ago and it's now available.

89:28Um and obviously if I hit the consumer view

89:31It just takes me right to it.

89:32Now the nice thing here is this is obviously not a great example because it's just me on this demo instance.

89:38I'm the only user.

89:40But if you had lots of people creating custom views, they'd come up as a list.

89:44Okay, so you start to get some insight into how the custom views are being used, what's in them, but also how many times they're being viewed.

89:52So if someone's custom view is maybe sort of rip-roaring away in terms of popularity

89:57Maybe it's worth building a custom workbook just for that particular section to add more context and richness to that particular analysis, seeing as that particular part of the business is more actively using the analytics tool.

90:09that you've offered them.

90:10So that's pretty much the feature in a nutshell.

90:12It's not really that um uh you know um sort of um complex um just looking at the comments the mean squared forecast

90:21Um I'm not entirely sure.

90:23That's a bit above my my my my understanding I think.

90:26I I need to go on Wikipedia that to check that

90:29Um they could have just queried the read-only Postgres repo before.

90:32Yes, they could have done.

90:33Uh they could have done.

90:34Um but as as you as you might not know, um with Tableau Online specifically, you don't actually have access to full access to the Postgres repo.

90:43It okay, a tableau online you only get two views in the custom admin view.

90:48In fact, um I can potentially uh show you that uh this is a slight tangent.

90:53If I go to the information lab demo

90:56I think I can actually show you the admin insights view if I just go to um explore.

91:03Admin insights.

91:04So this is this is the um

91:07This is the view that you get and you get these two data sources TS events and TS users.

91:11These are the only two views you get so you don't get the full access to the Postgres view.

91:15So custom views don't feature heavily in these

91:18uh uh sets but in tableau server absolutely you can do that um but for table online you can't so that's i think that's why they're really creating this because on table online where people are mostly living in this interface you don't have that option so that's uh

91:31Uh it's a little bit more of a trickier examples.

91:33Um great.

91:37So um

91:39If I go back to this list, block comments.

91:42How have I not shown this feature already?

91:43Man, I've had so many great opportunities to show you this, and I haven't.

91:46This is available on both desktop and web editor.

91:49it so I'm not gonna uh labor the point here I'm just gonna create a calculated field and I'm gonna make this larger so you can see so previously if you wanted to create a calculation you just do something

91:58something like this.

91:58So you just say summer cells and that's that hasn't changed.

92:02Obviously I can't type so um that doesn't help me.

92:06But if I can type properly, I I can just write a simple function.

92:09Then I might want to comment it

92:10And I'd do a double forward slash and I'd add my comment here.

92:15Now the problem is that if I went to a new line, I'd have to keep doing that.

92:19I'd have to keep creating this.

92:22Okay, not ideal.

92:24In 2020.

92:254 you can just do uh forward slash star and say what you want.

92:30uh carry on on a new line and then end it here like that and that's it you can do multi-line comments

92:39So what this enables is the ability to go ham on ASCII art if you really want to do that.

92:45You can uh oh god that's really buggy, let's not do that.

92:48Uh let's just paste this like this

92:51Let's paste this like this.

92:53I can give it a nice header.

92:55Yeah, you can get these back in back back when I was gosh when I think I was 15 or 16

93:01I used to go onto game forums and uh game forums used to write their walkthrough guides in text and this is exactly how they look.

93:09So they'll have like level one, uh do this, do that.

93:11You could get that into your calculation window now if you're really, really fond of that.

93:15Okay.

93:15So that's pretty much the feature in a nutshell.

93:18Block comments.

93:18It's really nice quality of life improvement for those people who are more verbose in terms of their comments, they now have a better way of doing that.

93:26It's also a nice way of commenting large sections or large calculations.

93:30that you thought worked and didn't work, that's actually probably what I'd use it for.

93:33Um, because I can just comment out a section of my calculation very quickly and then carry on with the rest of it.

93:37And then if I need it, I can comment it back in

93:40So that's actually the really useful thing there, being able to comment a section of your calculation uh in a in in a nutshell.

93:45Okay.

93:46So that's

93:47That's that.

93:48Let's go back to my list of features.

93:50So block comments.

93:51We've done.

93:52Okay, so we have a few data connectors.

93:54What I'm going to do is I'm just going to switch to data connectors.

93:57We've got another maybe 10-15 minutes before we're done.

93:59I appreciate I'm running over, but um I also like to cover things properly.

94:03So let's just touch on some of these.

94:05So

94:06The Salesforce Connector API got an update.

94:08Okay, so essentially what this means is that more Salesforce um objects are now visible when you connect to them in Tableau.

94:15If you ever use Tableau with Salesforce, you'll notice that certain types of fields don't come through from Salesforce.

94:21because there may be bespoke fields in Salesforce and therefore they don't show up in Tableau.

94:25But now what they're doing obviously because Salesforce acquired Tableau.

94:29You're now starting to see better integration between Tableau and Salesforce.

94:32And every single release, they've actually improved the quality of the connector and the way things work.

94:38Okay, so this is something just to be aware of if you're using Salesforce

94:41This is something to worth you know keeping a pulse on because you might have tried something literally last month and now this month it's possible.

94:47Okay.

94:48The next thing is the Data Arma Connector.

94:50Now this Data Arma Connector actually touches on two things, okay

94:53The Data Armour Connector is actually now available as part of the um uh let's see if I can find it here.

94:59The um it's not actually listed here, it's something called partner extensions.

95:04Partner extensions are essentially ways where partners of Tableau can write their own data connections.

95:09Okay.

95:09If I click on this link to the Tableau extensions gallery.

95:12and we just go there, you'll notice that I actually arrive at this page which is slightly different to what I've been used to and I'm actually able to download a connector here for Salesforce DataRama.

95:23So if I go back to the extension gallery

95:26you'll notice that the new thing they've done in 2020.

95:284 is they've actually split out the connectors gallery into its own space.

95:32Whereas before dashboard extensions had a dedicated place here.

95:36They now have an extension gallery for connectors.

95:38And so what people are doing is these partners are writing their own connectors for Tableau with their own systems.

95:46And DataRama and Salesforce are one of those people.

95:48Elastic are also another well-used

95:52uh connector that people really want because uh quite a few companies do actually use elastic.

95:57Uh especially startups, they they use that to

96:00Do things that are slightly sort of um more new age in terms of uh innovation and and and analytics.

96:05So all of these are now available.

96:07Now the key thing to bear in mind is that

96:09the partner has to build and maintain these and they don't have the traditional installation path as you used to.

96:15You see here I have to download a file

96:17Then I have to put it in a specific place on my computer, close tableau, open tableau, then you're able to see it.

96:23But it's a fairly it's a fairly straightforward process on a Mac, which is not really going to be that useful for many people.

96:28I'm just looking at yeah, I do have a screen this far to my left.

96:31I'm not

96:31trying to uh break my neck here.

96:33If I go to my documents and my tableau repository um and I open up my data sources um

96:39you will see actually no not my data sources it's actually connectors you'll see there's a connectors folder in here and what you need to do is actually download the file into this location okay I recently wiped my laptop so I actually cleared the connector

96:53that I previously had, but you see here that you need to put it inside of your Tableau repository.

96:57Then once you've done that, you need to install any driver that that tool might need in terms of

97:02um you know things like a JDBC connector if necessary or any other drivers in the back end.

97:07Then once you've done that you activate the connector and everything's pretty much

97:10much good to go.

97:11It's installed and ready to work.

97:13Okay.

97:13So that's something to be aware of.

97:15If you're looking for a connector, you can now lobby your technology partner to build their own connector for Tableau rather than waiting for Tableau to incorporate a slightly weird or quirky database they might

97:25never um come and come come come come to use so if I go back um uh I need to go back quite a few steps don't I go back to 2020.

97:344 new features

97:36Hopefully it takes me back to where I was on the page.

97:38No, it won't.

97:39So let's go back down.

97:41We were quite far down.

97:42I talked about Salesforce API and Data Arma Connector.

97:45Ask Stata, I'll come back to in a second

97:47Dynamic parameter improvements.

97:49This is a small quality of life feature.

97:51If I just go back in here, I close this calculation and create a parameter and let's make this a date.

97:59Okay, and I'll call this date parameter.

98:03And make it arrange.

98:04It needs to be arranged for this to work properly.

98:06Um, set values from order dates, um, and it goes and gets them.

98:10But obviously, I here I have an option to

98:12to choose what happens when the workbook opens and here I can go select order day okay so what now happens if I just go back to the setting is um you can now set a parameter range to fields min and max value upon opening the workbook or when the fields data source

98:25is refreshed.

98:26The actual quality of life um improvement here is just you've just got a few more options essentially okay you can choose um a few more things uh

98:35It's it's really sort of a light touch uh fix here.

98:39It's not like a a breakthrough feature.

98:41Actually, two versions back they added the ability for dynamic

98:43parameters so you could essentially get 80% of the value but what they've been doing is just improving the experience when you use it on web edit but also here on um tableau desktop okay

98:55If I go back, um Snowflake OAuth improvements, um that that's that's um that's a quite a you know specific feature.

99:03Um I don't use Snowflake so I can't really talk much this

99:06feature to talk about how much better the OAuth is.

99:09Analytics extensions for Tableau Online.

99:12So what you can now do is you can now enable um

99:16R and Python on your Tableau Online instance, okay?

99:18This requires you to do a couple of things in the back end and set things up.

99:23But once you've done that, your R or Python server obviously needs to be public to Tableau Online for it to work

99:29Um which can be s slightly sort of difficult because in some companies you don't have your tableau um sorry you don't have your R or Python server in a public place.

99:37You have it some

99:38somewhere internal.

99:39So that that can be a little bit of a cumbersome thing to set up just to make sure everything's working right.

99:44We touched on Tableau Server Management Improvement Features.

99:48Backups are faster essentially.

99:51There's a now an improved way of creating backups which is uh easier and gives better visibility of sort of what's going on.

99:57And if something fails you can retry from the last checkpoint so you don't have to start from scratch as well

100:01Which is a a really nice um uh sort of quality of life improvement.

100:05Um a couple of last few things.

100:08So we've got RS data to look back at.

100:10Data quality warning on metrics.

100:12This is useful.

100:13Uh data quality warnings are essentially on Tableau Online when

100:16uh you create one of those metrics what you didn't have a bit like in tableau prep earlier on where we saw we didn't see the data quality warnings you didn't see those in metrics either so now those are available

100:27Merging duplicate external assets and tableau catalog metrics and integration.

100:32Now these these two things are related to the data management add-on and specifically inside of the data management add-on is something

100:39called tableau catalog and with those two things you get these two features okay so again what they keep doing over the last I'd say six releases is keep on improving this lineage feature

100:48Because essentially what they're trying to do is make this a one-stop shop for handling and managing metadata, especially when you've got things like duplicate data connections.

100:57Um you've got two people connecting to the same thing, being able to merge them where it makes sense, they might be authenticating individually, and so maybe you don't want to do that, but where it makes sense you might actually want to merge these

101:08assets together.

101:09So the tableau talks to one thing and logs in once rather than logging in a hundred times because a hundred different people have uh have created the the same asset.

101:17Okay.

101:17And then again with Azure, I'm not exactly a um an Azure expert, but long story short, um what you can now do is you can now have an external repository in Azure rather than necessarily using um

101:31uh the one that was sort of built into Tableau.

101:34So these two are sort of very unique features for uh Microsoft Azure customers essentially.

101:40And then the last thing is multi multi multiple key activation on Tableau

101:44server so you can enter a bunch of keys and activate them all at once rather than having to do one at a time which was uh again if you're doing a server upgrade or install and you've got several keys to activate it can be a little bit of a painful process

101:56Okay, so the last thing to do is just show you this Ask Data Improvements option here.

102:01So let's go back to uh do I actually have my other browser open?

102:05I think I do.

102:06Here we go.

102:07And uh if I go back to explore

102:10Go to default and let me look at this as a list again and let's find my demo output.

102:16This is my data source that I created.

102:19at the very beginning of this live stream and the thing you can see here is that I'm loading straight into the RSData interface.

102:26Now Rs Data, a lot of people don't like it.

102:28I'm actually starting to become quite fond of it because it's starting to work a lot more

102:32So what I'm gonna do is I'm just gonna ask a very basic thing.

102:35Um profit um profit broken down by category

102:44Um uh by month.

102:47Okay, so I just typed that, sort of didn't really look at what I was typing.

102:51I've hit enter, nothing's happened.

102:53Uh let's try that again.

102:54I think I must have missed the uh the key

102:58And let's have a look.

102:59So, yeah, it's show me the profit by category, uh, and it's broken it uh down across months.

103:05So it's built exactly what I wanted to see, which is great.

103:07Um the nice thing is that in this version they've added a stacked bar chart which is this, so you can see this in a slightly different way, which is quite nice.

103:15So you can actually type as a stacked spar chart if you wanted to.

103:18I'm not going to do that because I actually quite liked the line before.

103:22Now, what I can also do is I can also type in a date range.

103:25So I can say between uh March 2006 and

103:29uh feb 2017 okay let's see if that works I just hit enter march 2006 and february 2017 and let's click on this to actually see what's going on

103:39Okay, so March 2006, there wasn't any data.

103:42I think my data set actually starts in 2017.

103:45So let me try that again and be a little bit more uh brutal.

103:49Let's say between July 2017 and March 2019

103:52Okay, so between uh March twenty seventeen and july twenty nineteen, just hit enter.

104:01And yeah, it's pretty much filtered that quite nicely.

104:03But the nice thing here is that obviously if you do a date range, it's actually giving you this nice interface as well.

104:09So you can play around with this.

104:10So the way you have to think about this is that essentially if you think about um you know new users to Tableau, completely brand new, never used Tableau before.

104:19If the first thing they have to do is learn Tableau to get to it and start using it, that's actually not an ideal experience.

104:24So if they can come to something like this and just simply type an art question, an answer, and get it

104:30back.

104:31That's really really valuable because now people are able to build something and if they're learning, they can obviously open this up, they can save this, save it as a new asset.

104:39I'll just call this as a demo worksheet even.

104:43Hit save.

104:45And it opens it up and they've got a chart straight away.

104:48Um, what you then want is to give this sheet a home and finish it off.

104:52You can go in here and edit it, and you can start to play with it and understand how it worked.

104:56But you have a nicely really easy to use way of getting involved.

105:00Now, obviously this is not ideal.

105:03This is going to improve over time.

105:04The natural language

105:06understanding is going to improve.

105:07Um admins have to do quite a bit of work adding things like synonyms in and data roles.

105:13We've talked about that earlier on.

105:14they have to be added in here.

105:15There's a whole sort of uh interface for for letting you do that here um whether it's things like date time filters um whether it's in individual fields you can go in add synonyms you can also uh

105:27control the data roles and then publish them and then those data roles can be applied time and time again.

105:32So there's a little bit of handiwork that needs to be set up and you've also got the complexity of language because everyone has to speak English for this to work and that doesn't always happen.

105:42So there's a little bit of overhead, but this feature is getting really really good.

105:46A while back I I wasn't able to type and get a sort of a smooth answer like I just did.

105:50It's a lot smoother and it's a lot lot

105:52lot faster to work with.

105:54Okay.

105:55So we're finally there.

105:57Gosh, that that took a lot longer than I was hoping.

105:59I'm really sorry to keep anyway.

106:00If if you've watched this entire stream, I'd love you to leave a comment.

106:03Oma, I know you've watched the entire stream

106:06You've been very, very engaged throughout the session.

106:08Thank you for spending two hours, nearly two hours with us today on an evening.

106:13I don't know where in the world you are.

106:15Let me know where you are in the world.

106:17uh I'd love to sort of be able to um know where everyone is uh joining in from.

106:22Um just a few things before we finish off

106:25What I said earlier on is I was going to show you a new feature that is currently in preview.

106:30And this is actually something called collections.

106:32Let me just show you this.

106:33So Tableau Collections.

106:36Okay, if I just search this, you actually find um that this comes up here okay in the documentation, which it's actually live in 2020.

106:444.

106:45And so they've got a really nice instruction guide here on how collections work.

106:48So this is something that anyone can go and look at straight away.

106:51It's public and what I've actually been able to do is secure access to the preview.

106:57So

106:57We can actually see what that looks like.

106:58So earlier on I talked about something unique.

107:01This is uh this is essentially that.

107:03So what I did earlier on, I'm not gonna sort of break this down too much.

107:06I think it's unfair to um you know critique a feature that's still in development and they're asking for feedback on

107:11So if you'd like to give feedback, you can go to that page I showed you earlier on, apply to be into the in the preview so you can try collections for yourself.

107:21I've created a collection here.

107:22I did this yesterday.

107:23It's really, really nice.

107:24And you can actually see here that I've actually put a bunch of things into this.

107:28And so the way collections work is essentially a way of

107:31creating your own little project folder which can store pretty much any asset that you want whether it's a flow, a data source, a metric, a workbook or a worksheet.

107:41Everything can come into one place.

107:43Now on conference I actually demoed something more advanced than this, which let me create some sort of a feature page as it were with metrics, a dashboard, and things I could actually interact with to analyze and filter data with.

107:55in one place.

107:56So I'm hoping that comes in the near future.

107:57But as a starting point, I can actually see quite a few use cases for this because each of these collections have a link.

108:03you can share them uh as you can do with most things uh in tableau they have a new sort of um icon which shows that they're here to stay tableau invest into some sort of new parts of their interface

108:13they really do it properly.

108:15I can add, I can invite usernames and people to this.

108:18They're really easy to create.

108:19Essentially you just need to go to any content.

108:21So if I go to this overview for example

108:23I just have this option to add to collection.

108:25When I go to that, it gives me the option to add it to my existing collection.

108:30Or I can create a new one.

108:31So I'll call this new collection.

108:332020.

108:354 demo and uh create.

108:37Okay, an unexpected error happened.

108:40I think this is because I had a point in there, so let's try creating that.

108:43Yeah, computers tend to hate points, okay?

108:46So um it's interesting because it seems to have duplicated it.

108:49But again I said this earlier on.

108:51This is this is still in preview, so it will be buggy.

108:53We shouldn't be judging this based on um how it works and you can see these these are private.

108:58So the one that it created, let me delete one of them.

109:01I hope this is the one with the thing uh thing I just saved

109:04uh if not uh be a little bit stuck yeah this is the one here on the left hand side so you can see I added this item to this collection I can obviously change um the description

109:16And I can also change the name uh if I just go to these three dots and rename.

109:20I can call yeah, I actually already named it something sensor.

109:24So if you rename it, you rename it.

109:26The other thing is you can change the privacy.

109:28So you can go to permissions here and you can make it public or private.

109:31But this is a very, very early stab at this.

109:34So

109:35Tableau really want feedback on this.

109:36I personally quite like it.

109:38Any capability where I can essentially customize the way I use Tableau is something very welcome because Tableau server can be a very overwhelming place.

109:47So if you can create a bunch of assets that are important for someone and say, look, here's a collection for new joiners on Tableau Server, right?

109:56It can walk them through the experience of using Tableau Server, give them activities all in one place, and it's really nice and easy to work with.

110:03Then as they get more advanced, they can create their own collections.

110:06As you work as a team, you can create a collection of items with

110:08assets that you've pulled from different projects across Tableau Server into one place.

110:13I obviously want more organization of that.

110:15So I'd like to be able to have a data section, a flow section, a metric section, a little bit more context.

110:22would would be very very useful okay so that's pretty much it uh let me head over back now to the question I'm gonna actually drag this on screen so we can actually address this uh very easily um

110:33Right, uh so I'm I see you you entered a block comment in the comments.

110:38That's that's really uh that's really good.

110:41Uh great session.

110:42Uh got to go, it's 9.

110:4330 over here.

110:44Oh, okay.

110:44Sorry we lost you there.

110:45Um

110:46Thank you very much for nominating me.

110:49God, I'm a little bit uh I never know what to say when uh uh people sort of speak positively you, but um I I really appreciate I really appreciate um when people

110:58I take the time to watch the content, let alone um, you know, is uh share some kind words.

111:02So how do you add them to online?

111:04Oh, okay.

111:05So um

111:06Omar, if you've gone, you're probably not seeing this, but essentially if you go to the uh very top of this collection page, you can sign up for the preview here, okay?

111:15So it's not something that comes on by default.

111:17When you uh simply put your name, email and the URL for the Tableau Online site in here.

111:22Uh it can't be a dev instance of Tableau Online, so it can't be one of the beta instances like I use for this demo.

111:29has to be an actual functional Tableau Online instance because it's being switched on in a slightly different way so it can't be on this on a test instance.

111:36So that's one quirk to be aware of.

111:38Head over to this page and you basically end up getting into the preview

111:42And it's nothing sort of hidden.

111:43It's very uh publicized here at the very top of the page.

111:46But you have to know about it to find out about it if that makes sense.

111:49So that's that's how that works.

111:51Um and does adding them make um make them available for prep as well?

111:55So uh yeah, so the collections support everything.

111:58So you could have a collection which is just flows, you could have a collection which

112:02is flows and their data sets.

112:03It's uh it's something that cuts across the entire server.

112:06I think that's what you're asking.

112:08So collections work across the whole entire environment.

112:11Can you show the merge duplicate assets on catalog?

112:14I would want to show that except for

112:17On this particular um instance, let me go to external assets.

112:24It's a little bit difficult because I need to find uh a good example.

112:28So what I'll probably do, I won't do it here on stream because we're running a little bit long, but what I will do is I'll create a separate video and I'll upload it to the channel uh probably later this week, if not early next week.

112:38So just if you're looking for that, check out this channel.

112:41I'll show you how to do that specifically as a follow-on from this live stream.

112:44And then I'll be able to create like an actual example where you can see that work properly if that

112:50makes sense okay so that's pretty much all the comments um thank you everyone who's joined in um if i just close this uh chat tab now um i'd laugh right now if there if people uh hadn't been watching this whole time because something happened all

113:03something disconnected uh but thankfully everything's still working so thank you very much everyone who's joined i really really appreciate it we spent two hours together um it's been amazing um i think maybe next time we should spend

113:15These up, maybe not.

113:16I actually quite liked doing it in one go, it gives it a nice holistic feel.

113:20Um, definitely be sure to check out our meetup page, and this is a great place to find out all the future events that we've going.

113:26Future live streams will also be here.

113:28Um

113:29Be sure to check out our YouTube channel.

113:31This is obviously where this live stream is going to go.

113:33You can see that it's uh live there already.

113:36And um, yeah.

113:37Just really engage with us.

113:39Um, we appreciate that it's you know it's a really tough time for a lot of people around the world.

113:43And so anything we can do, especially when it comes to Tableau and Ultra.

113:46to help you uh with your experience of it anything we can do to help um you know let us know in the comments let us know through our events be in touch with us on social media

113:54And uh wherever you are in the world, I hope I hope um I hope things are getting going well.

113:58And um yeah, we'll catch you in the next live stream.

114:01Thanks, thanks for watching, thanks for tuning in.