 Hello everyone, welcome to this UK Data Service Workshop on Geographical Data Visualisation of UK Census Data. Your presenter today will be James Cron of the UK Data Service. He's based at Adina at the University of Edinburgh. And now over to James. Thanks Jill. The layout of the day is going to be a few presentations from me and quite a few actual hands-on practical exercises. The emphasis in the workshop is to actually get your hands on using data and software. Let's continue. So we'll start with a quick introduction to the UK Census, Census data and how Census data can be visualised geographically. So what is the Census? Well the Census is a survey of the UK population held every 10 years by the devolved national statistical agencies. So in England that's Officer National Statistics, in Scotland National Archives of Scotland and in Ireland it's Nizra. In England it was taken place in 2021, but in Scotland because of the Covid pandemic it's only happening this year. And in fact in Scotland Census Day 2022 is on this coming Sunday. So you're in Scotland you've not actually filled your Census form and yet do it by Sunday. And the population fills in a Census form and answers questions relating to them either as an individual or the household which they reside in. And you can see on the slide the household questions and on the right the individual questions. And it's a way of the government getting information on the population every 10 years. So it's a great source of information. And what the Census agencies do once they've gathered all this information is that they process it and create various output Census data. And they censor aggregate the data. So the data comes out as counts of people or households for particular socioeconomic characteristics. So what you can see in the slide here is an area of Edinburgh. The Census data comes from various small areas of geographies. And within those small areas you get different types of data. So for example you get the total population, number of households and the number of people in that area who are male or female as well as information such as detached houses or flats. And the Census data itself is a variable at various different output geographies. So you can get data for all of the UK and then it will break down into countries, so England, Scotland, Wales or local authorities. So Manchester University Authority for example. And then wards, electoral areas and these tend to nest in some cases. Within the UK data service we provide a one-stop shop to access social science data and this includes access to Census data as part of our remit. So for the UK data service you can access data from 1971 through to 2011. We expect the latest Census data, the 2021 stuff for England to be made available from 2020 onwards possibly for applications. The Scottish stuff will obviously be different later because they've not actually done the 22 Census yet in Scotland. And within the UK data service we have different applications to access different types of Census data. So the main Census data is a Census aggregate data and those are like the grouped population stats. And then we have Census microdata which provides quite detailed information for smaller numbers of people and Census flow data which provides data on migration. So that's how they travel to work on Census Day and also their last address before the Census. So it gives you some information on how people have been moving around the country. And then we also have supporting data such as Census boundary data and these are the geographic boundaries for the various Census output geographies which can be used to visualize the Census data. And in this workshop today we'll be using Census boundary data in association with the Census stats to create some maps and other types of visualization. So today you'll get access, you'll be using both Census aggregate data and Census boundary data. So the purpose for this first exercise in this first part of the workshop is to visualize Census data using choropleth maps. So if we have tables of Census districts we can join or link them to Census boundaries like on the slide. So at the top left we have a table of Census statistics and this shows information from the Census in this case for local authorities. So we have Edinburgh, Brentfisher and West Lovian. And each of these areas has a unique geographical identifier which is a line digit code that uniquely identifies that record within the Census data. And you can see the information such as some mails and the manufacturing number of mails employed in manufacturing. And we can have the same geography in terms of the actual boundaries. And these describe that geography on the ground. So the extent of say Edinburgh or East and West Lovian. And by linking the Census stats to the boundaries we then have some way we can then describe the Census stats as a map. So in this case we can say that Edinburgh has been assigned the value of 5.27% of mails are employed in manufacturing. Whereas in West Lovian 14.93% of mails are employed in manufacturing. But it's not very interesting just like this because we can only look at the different values to see where the different publishing stats are. It's far better if we can colorize those regions. And this is what our coropleth map does. It basically shades the polyons of the Census areas according to the Census variable which is linked to those boundaries. So for example high Census stat values could have like high shades of green and low values of lighter shades. And that way when you view the entire dataset you can quite easily see where the high areas of the particular stat are and the low areas are. And that gives you a great way of like getting some a look into the actual data. So for example this map shows the percentage of people employed in agriculture in forestry and fishery. And you can quite clearly see the difference in the urban and rural split. And also the difference in the Scottish Highlands and the Scottish borders in terms of the number of people employed in the fishery and forestry. Now these are a type of coropleth map with a univariate coropleth map in which it only displays a single variable. You also get more complicated coropleth maps which attempts to display two variables at once. And that could be quite interesting because it means you can display multiple variables at the same time. So here we have the first coropleth map which is showing poor health from low to high and a second coropleth map which is unemployment from low to high. And we can basically combine those together to create a map that shows the two variables together. So on the bivariate map the very pale areas of both unemployment and poor health is low with the dark areas where unemployment and health is high. So it's a nice way of instead of showing two separate coropleth maps you can just show a single one. So when you're creating coropleth maps there's some quite this specific guidance which can help you in this process. So the thing is to choose the correct output census geography. You remember I said that the census variables are available at different levels of geography. You have to standardize your census variables. You have to choose an appropriate classification method and then choose the color ramp to tell you how to actually describe the data. So in choosing the output census geography you have to be picked what you think is most appropriate for your data. So for example we could display the same variable by local authority, by medium super output areas or by output areas. And by mapping the analyzer data at different levels you can get different insights. So for example patterns present at one scale may not be displayed at our scales because at the larger areas the data will pretend to get smoothed whereas in much more high resolution data you might get a lot more noise. So the local authority is good for picking up general patterns where the output area level is a lot more detailed and allows you to drill into detail in much finer detail. In terms of standardizing census variables by themselves the census data gives you raw counts so the number of people per output area or the number of people per county. And you can't just map it like that as it is because the high counts will simply show you where the people are. Instead you've got to do some sort of standardization so that you can compare that area to another part of the country in a standard way. There are two ways you can do that. You can divide that count by the area of the actual area of the boundary or more likely you can just you can divide it by the total population size of that area. So instead of showing a raw count you're showing proportion. Instead of saying in this census area there are 500 people you're saying that in this census there 15% of those people are employed in manufacturing or male or female or travels work by car or bicycle and that because it's a standard measure you can compare other areas. And then when we construct a chlorophyll map we also need to perform classification. So our raw data might have values from 3 to 93 but because the way is the minute we have unique values what we need to do is like generalize the data into data classes so that we can like try and find the patterns within it. So we might have five classes here say from 3 to 20, 21 to 38, 39 to 56, 57 to 74 or 75 plus. And what we do when we apply classes is we decide how the data is assigned to each of these classes and there are different classification methods available. Some of it you'll see in the QJS exercise which follows this talk. So for example this quantile, equal interval, natural breaks or manual. And the classification maybe we choose will cause the data to be drawn in particular ways. So for example I have the same data set here and depending on the number of classes I used to classify the data and the classification method I use I can get various, I can get different sorts of maps being shown. And the point is that no classification method is right or wrong but the classification method you do choose should be based on the characteristics of the data to avoid constructing misleading maps. And then once you've done applied the classification we then need to apply a color ramp which basically tells you how the polygons are shaded and different color ramps are applied to different sorts of data. So sequential would simply be going from a low value to a high value. And you can see there's a graduation from light to dark. Diverging would be data plus or minus around a central zero value. So that could be areas that are increasing value versus decreasing value. Whereas qualitative would be non-numeric data. So this could be data that's showing like a, it could be a land cover or it could just be like you could have an urban rural classification where some polygons are describing urban areas and some rural areas. And then in that case it wouldn't be assessed to describe it quantitatively. You would simply describe have a different color which is a categorical thing. So you'll get to practice all these creating chlorophyll maps in the curious exercise that follows. But to show you what's going on I will first run through the exercise live and then you'll have time to do it yourself. So this is the UK data service and I say as part of this you provide census data. And the part of the UK data service that we're going to obtain census data for in order to create a chlorophyll map from is called infuse. Are we going to infuse provides access both to 2011 census data and 2001 census data? Are we going to grab data for 2011? So what you would do in the exercise that follows this is you would go to infuse and click the 2011 census data box here. And when you're using infuse there are two means of like making your first selection of the data. You can either select by geography or by topics. Are we going to select by geography first? So infuse is doing a lot of work in the background so there may be some delays as we run through the using our service. Okay so because we selected want to go by geography we've got the various geographies that are available. So I'm going to create a map for all of the UK and I'm going to obtain census data by a local authority. So for each of England, Northern Ireland, Scotland and Wales I'm going to select local authorities like so. So I select those and then I add them to my I guess my shopping basket almost. And then what infuse has done is gone away and worked out what data is available at these geographies and you can see there's like a load of different combinations of data here. And using these filters at left I can constrain the sort of data that I want to I can constrain I can limit what data I actually want. So in my case I can apply multiple filters so let's see I can constrain by sex and then I can also constrain by industry. And you can see it's dropped a number of combinations down from about 80 just to a single topic combination. And if I select that topic combination I get some information about what's available. So you can spend your time reading through this and within that topic I now have to select the actual variables. So I'm interested in creating a UK map of I think the percentage of males and females employing manufacturing. So the variables I'm going to select are age 16 to 74. I want total industry and manufacturing. I need total industry because I want to use that as a when I'm doing my standardization of manufacturing so that I can show my manufacturing as a percentage of a total employment and then I want both males and females. So you see I grabbed four variables here and so infuse is ready. So at this point I can click get to the data and infuse will go away and pull that data out for me and let me just download it. So it comes as a zip file which I can extract and within the zip file you basically get three individual files. You get citations which I think just describe provides copyright statements you have to use if you want to use the data. You get meta which describes the actual variables and I guess the units to use and the data itself is this CSV file called data age equate intersect units and if I open this in library office so if you're using windows you can use excel. You can see what the data looks like that's come from infuse. Before I use this in QGIS I need to do some cleaning up and tidying up all the data. So the first thing I do you can see infuse has created this strange line here where there's like null entries so I'm just going to get rid of this line and then I'll tidy up some of the columns and I'm just going to rename these columns here. So I just have to refer to my notes so so this column is the total number of males within this local authority. This column is the total number of females and then these columns are the males or females employed in manufacturing so I'll just rename them to something more memorable male manufacturing and female manufacturer. And what we want to do now we want to create we want to standardize our data so rather than in our Mac QGIS just showing this raw male manufacturing value 192 you want to show that as a percentage of all males so we have to do like a simple calculation in our spreadsheet here. So wrong column that's more likely okay and let me just give this a name okay I mean this similar thing for the females yep and then I can apply that to all the cells and to likewise for female. Now I just save my data now and I can check it what it looks like in the actual CSV and you can see that the numeric data looks fine we have this problem with some of the the geocode which in our case the geographic identifier and you can see the way that infuses worked it's inserted trailing spaces to the northern island geo IDs so we try using QGIS it'll create problems so we have to go back into the data in Excel and correct these. So what I'm going to do is I'm just going to simply trim these northern island geo IDs to strip off that at that white space I need to apply this to the first 27 rows I'll just copy that and paste it back into position over here okay sorry about that don't often use Excel for data analysis so hopefully that's okay let me just check it in a word pad yes that looks correct great so we grabbed data from infuse and we've done some preparation on it so it's now ready for us to be able to map being QGIS so we got the CSV data the census data we now need to get the actual boundaries that we're going to create our map from and that's a little from a different part of the UK data service called um let's look here so UK data service and I scroll down to the bottom here in fact find data you want to browse and access data I think yep I have to scroll to the bottom here to find uh browse open census data yep you want census boundary data um this will take us to boundary data selector what what we actually want is easy download so we want to grab some data that matches to infuse so we go to this tab here and we want to grab some local authorities this is all explained in the PDF you'll use in a bit and you want to download the features in shape power format so if I look at the downloaded data and let me just open in QGIS so QGIS is a desktop GIS um and you've never used the GIS before it might seem overwhelming but QGIS um but the notes we've given you will try and help you use it the first thing you have to do when you use QGIS is to um geographic data is provided in different reference systems because it has to describe different parts of the world um by default QGIS will default to showing data in lat long and using a global projection so we first have to change this so they use the British national grid which all data within the UK data service is provided in so this is explained in your notes so you basically you'll go to the bottom here click this button and pick British national grid and apply an okay and you can see that it's changed from 4326 to 2700 and then we can now add that vector shape file which will be downloaded from the UK data service so this button here is called the data source manager and this is the button you always use to add data to QGIS it could be a shape file or a CSV file so I'll open the data source manager got different at the left here different types of data you can add so I want to add vector data so I click vector and I use the two arrows here to choose my data set so it was that infused this layer clip so I just open add and you can see these are the boundaries that have been added to QGIS from the UK data service and we now to add our CSV file so again we use the data source manager so back into that folder select our data CSV and you can see that QGIS provides a preview of what that data looks like at the bottom here and you can see it's got our new columns we created in excel or library office when you do this you have to make sure that under geometry definition it's set to no geometry because otherwise QGIS will try and create like a point and it won't find any coordinates so that won't work properly so we just add and then close and we can open that data there so you can see we've got the same data we had from excel or library office in QGIS and what we now have to do is join that data of of census statistics to the census boundaries and we can do this by using a thing called join so we would double click the layer go to joins at left and click we want to create a new join so click the add green button because I only got one CSV file loaded it's been added automatically to the top box here if you add more than one you have to pick which one to use but because we only have one we only have one choice we now have to specify which field within the CSV file contains the geographic identifier and which field within the boundaries contains the geographic identifier which QGIS will use to relate the two tables so in the case of us in the geocode in the CSV data and again the geocode in the boundaries so we apply that and now if I open the actual table of the shapefile you can see that our data from the census stats has been joined to the boundaries and in fact I can click on any of the polygons and I can see the proportion of males or females within that area and so what we now want to do is create a crawl path map using our now census stats joined to the boundaries and to do this double click from left choose symbology are from top pick graduated I mean I have to show we have to pick which variable we want to display as our crawl path map so let's say I want to show the proportion of males employed in manufacturing and then we click the classify button and like I was saying in the presentation we have to we have to pick the classification we have to want to use so I'm going to go for natural break genks and as I said in the slides you also have to decide what color amp to apply so I'm going to show them in green from low to high value and click the apply button and we have a crawl path map of our data so so low values of green or white or with this low proportion of the male population employed in manufacturing dark areas or with it's a lot higher so you can pick out around the areas of Newport and Wales which is like the steel mills the north of England sort of manufacturing and you'll see a lot lower in London where there's people who tend to be more employed in service sectors and within QGIS loads of other stuff you can do in the document you can actually create a print map layout which the PFO describe so I'm going to stop demoing QGIS for now and hand over to you for you to go through the workbook until three o'clock when I then we'll talk again about using cartograms and I've got another short presentation where I'm going to talk about creating cartograms and flow maps and after this there'll be a further QGIS exercise where you'll be able to create a cartogram so other than chlorophyll maps two other forms of geographical visualization commonly made using sensitive data are cartograms and flow maps so a cartogram is a special form of map projection where the polygons are drawn in proportion to the variable being mapped rather than the land area of the polygon so you can see on the top here we've got three different types of cartogram if a non-contiguous cartogram a contiguous cartogram on a darling cartogram and you can see that the the boundaries have been distorted according to an actual variable rather than just being based on the geographic extent I mean cartograms are great for like excuse me I'm just gonna have to cartograms are great because they mean if you have a chlorophyll map of Scotland you could have data shown for glas and reddenburg which occupy quite small areas and they tend to get lost within the bigger of the country areas they tend to get hidden and you might draw false conclusions from the map but using our cartogram we would just start the polygons based on the variable you get a you can see different patterns which is really quite interesting and there's sort of different types of cartogram and cartgrams feature quite extensively in the wild so to speak they were shown quite a lot in the guardian as part of their brexit reporting so you get a cartgram like this and you can see that although Scotland occupies quite a large geographic area that's not people living there it's quite small whereas London occupies a relatively sorry yeah so London has a lot more people living there than a much smaller area so when you distort the polygons by the population they tend to swell up and cartograms there's an entire book written by Danny Darling who's like a prominent census person which entirely uses cartograms rather than chlorophyll maps to depict census data it's a really cool book there's like each chapter goes through a different sort of type of census data and it's full of these wonderful cartgrams in the QJS exercise that follows this you'll have the opportunity to create your own cartogram what you'll be creating in the QJS exercise is a contiguous cartogram of this sort of like on the right here where you'll distort the boundaries and you'll basically yeah you'll distort the polygons according to some variable which is quite cool so another type of census data is census flow data and what census flow data is based on a question the census form which asks people for the address their place of work so what you end up with is a data like this or table here so down the left hand side you have what we got one two three four seven different regions and along the top you have the same seven and it basically depicts the migration of flow between each of the polygons the region sorry so since they're Northumberland County Durham Northumberland Newcastle Bantai Northumberland Gatehead and this is quite useful for planning purposes and this data is available through the wicked part of the UK data service and the sort of visitations you can create from that data tend to be spent these flow maps so what I've got is an example on the left here of the UK and this is showing the migration I think it's not migration it's the place of work between Edinburgh and the rest of the UK this was probably pre-pandemic so you can see this large flows from like Edinburgh down to London which is probably the people who tend to like commute from Edinburgh to London on a sort of Monday and Friday to work in the city of London and they go back to the Edinburgh at the end of the week it's the simple sort of diagram here you can create in QGIS and people have created more whizzy sort of type flow maps using sort of quick cool styling stuff alternative is it's the notch with the boundaries this all and creating some kind of chord diagrams which basically just like they map an origin to a destination I think this is like for the US so it's showing the flows from say New York to California and the width of the actual what we call chords it depicts how much flow there actually is and that's like an interesting way of showing flow data about using a map in some ways it'll be more powerful so we have an exercise now in QGIS where you'll create a chord diagram and again I'll do a first run through the exercise on my own machine and then we'll give you 20 minutes to do it yourself so let me just again we do some screen swapping so bear with me again and for this exercise we're not going to download data from the web we're going to use data that's been prepared beforehand and that is available in the zip file which you've been sent a link to I'm going to provide a link to afterwards so just bear with me I've already downloaded that document in the zip file into my desktop so if you've downloaded that data you'll see the contents you've got the workbooks and there's the census trading data folder within that census training data folder or four different subfolders the cartogram corpus and tableau we'll be using the tableau one later on as well so for now the cartogram one contains census data for leads at the middle super output error level and it contains a shape file plus a csd file so like we did in the chlorophyll exercise we'll open the shapefiling qjs and then we'll add the csd file and join the two together and then we'll create some visualizations so let me just open qjs first again so qjs is open just check again we have to set the projection to the British national grid same drill as before use the data connector to add the shapefile this time from the pre-downloaded data so here we've got middle layer super output areas for leads and just add the csd file again from the pre-downloaded data checking we've got no geometry defined and again same drill as before we have to join the data so this is all explained in the workbook so I can just check the attribute table and you can see we've joined the census stats to the boundaries this data actually shows it's tenure by household so it basically displays what people are privately renting which people have mortgages or which people are renting in local authorities and stuff and we can do what we did before and create a chlorophyll map to start with so I want to show the percentage by private renting let's go over my natural brakes again and green and you see the problem we've got like our small polygons in the city centers which we tend to get mixed by the outer polygons we want to actually create a cartogram qjs by default doesn't include the cartogram functionality but you can use a plugin to do it for us a plugin is just a piece of software that you add to qjs to do add extra functionality so we have to install it first and when it installs it as new buttons the qjs toolbar so it's listed just in case of picking that later and then you pick the variable you want to pick the cartogram from and it will go away and build the cartogram and then so there's our cartogram and there's our chlorophyll map and you can compare some people don't like cartograms because they think it distorts geography too much and you can't tell what's what but I think they're a good way of alternative way of viewing the data so I'll stop sharing now and you've got 20 minutes to work through the cartogram pdf and try and create a cartogram and again any problems put something in the question answer and we'll try and help you out I'll just show you the other screen so I can the links to the data workbooks so basically if you download that zip file we sent you earlier you should have that folder with a workbook pdf and that data folder with the leads data so it's just a case of working through that workbook and you shouldn't end up creating a cartogram the workbook is also available via this link and as I said at 3.30 we're going to take a 15 minute break so at 3.45 I'll start talking again and we'll move on to using Tableau or having QGIS so again any questions just pop them in the chat hello everyone so hopefully you managed to create a cartogram we're now going to start looking at um Tableau so this session's a bit different there's no actual slides as such we get it's a practical hands-on exercise just get it to drive straight using Tableau because we have limited time so again what I'm going to do is I'm going to show you what's going to happen in the exercise and then I'll give you time to go away and do it and again stick your questions and answers in the box so let me just do a screen shuffle again I think what Tableau is it only runs on Windows and Max it doesn't actually run on Linux which I'm using so I'm going to have to run Tableau instead of virtual machine and Windows on my Linux box so this is why you have this slightly strange setup here but I'll continue so the exercise actually has two parts to it is a first part where you basically map some global some global data by country and there's a second part where you map from the census data um and what we're going to do I suggest is we just do the second part using the census data so let me just try and remember how to do this process okay so first of all and we're going to use data inside that census training data folder as a Tableau sub folder which contains all the data you need so you want to do any more downloads so let me just fire up Tableau you see Tableau is a lot more slick than Secuges it's quite a nice piece of software and what I want to do we're going to be using two files we've been using this travel gm.csv file which might just open in something notepad and you can see this is data about travel to work by some geography which is wards so we've got a shape file for that so let me just extract that and these are wards for 2011 so a bit like in QGS we first have to add the csv data to Tableau and then the shape file to Tableau and then do a bit of joining to create stick put together and then do some manipulation of the data and then create some sort of graphics and visualization using Tableau so let me have a look so the first thing I'm going to do is connect to the csv file so this is travel gm.csv and Tableau will read the data you can get rid of this thing on the side here so this is our data in Tableau this data interpreter thing here Tableau will try and do some cleaning on the data what we want to do is grab all this data here and we want to basically pivot the data let me start again okay let's try that again the instructions are really clear so if you follow which I wasn't doing you should be okay so the data been pivoted so this means instead of like I say 30 rows with five columns we've now got 150 rows of a single column so it's just a different way of representing the data but we need to do that for the Tableau to create visualization from it and we might want to it's quite good in Tableau where you rename the columns so this is actually word and this is local authority and this is travel to work method and this is a percentage of people using that travel to work method okay so we've done some work on the data on the csv data rather and we now want to add the boundaries so in the same way you can add and connect to a spatial file which will be our shapefile and you can see that we've got these connections to the csv and the shapefile and we've also got the file here I mean now we should relate the csv to the shapefile so the first way to do this is to double click but this is called the Tableau canvas so we double click the csv file on the canvas it creates this and then we can try and connect the shapefile so I drag it across here and you can see that Tableau has been quite clever and it's trying to create a join by itself about us how to tell it but it doesn't know how to join the data so like in QGIS we have to specify which fields to use to do the join so it's geocode in the csv file and geocode one in the shapefile and you can see it's made a join because to our csv file in Tableau we have our attributes and we have a geometry column from that shapefile and now we can actually having prepped the data from Tableau we can now do the actually interesting stuff and create visualizations from it so Tableau has different types of there's different parts to Tableau there's the preparation bit here you can create worksheets which are like where you build traffic and plots and stuff there are dashboards where you can create like visualizations for consumption by end users this can contain multiple worksheets so you could have a dashboard showing a map or a graphic and then you could like publish that to Tableau Public Online and then someone could do your visualization on a website and there's this thing called Tableau Stories where you can basically create 10 different steps or so and allow users to page through different visualizations to tell a story about your data which can be quite nice but you basically start with worksheets you build your visualization worksheets and then you can add them to dashboards and stories so we want to create our first worksheet and Tableau is all drag and drop you basically have your tables your data and then your your work area and you just drag stuff across to it in order to create your visualizations so the first thing we want to do is just create a table of our data so I'm going to drag the travel work myth across to local authority and wards so we've got the outline of the table you don't actually have the values yet because you have to tell Tableau how to display them so you want to display by percentage of people so we have to to drag this across here yeah okay you have to double click rather than dragging so important to tip there double click don't drag so now we have a table which shows the data from our csv by ward and then local authority and we want to now create a bar chart so that's another type of sheet drag percentage of people columns and travel to work method yeah so we're now we're symbolizing the data broken down as bar charts by travel work method so you can see for harper green for example the travel work method by non-employment okay yeah driving a car a van it's 35 percent and then you get the different breakdowns so that's non-geographic we can because of the boundaries you can create a map so again we create another sheet and we can add the geometry by double clicking and you can see we've got our senses boundaries within table and the minute we mouse over we just see the we don't we don't have to click on the individual areas but we can correct this grabbing words and now we get the individual areas but there's still some strange things going on because we've got like multiple things showing particular regions which we can correct so we can grab local authority as well so it disagrees by local authority as well i think the problem before was you could have awards with the same name in multiple local authorities so when it's grouping it's not showing the individual ones and then we can create a dashboard from our two plots from our bar chart and our map by clicking the new dashboard button so if you hover over here you can see the types of things you created so we've created a map so we can drag that across create a bar chart and what's nice about tableau is you can use one visualization to filter the other so you can use this as a filter and select asli bridge and it will then zoom to the boundary and you can see the stack of the boundaries or we can do the opposite we can use the map as a filter for the bar chart so and then we can zoom into one of the polygons click down here and it will show the bar chart just for that polygon so this is the exercise that you will now do and i suggest you start with the second step but if say if you get through that and you've got time left then you can do the first one as well it's a different exercise there's different data so you're not going to cause any problems so let me just swap my screen again if you download the census data pack there's a table intro workbook in it and again your data is within that folder you'll have a census training data folder and a table sub folder with all the data you need so just work for the pdf and any questions in the chat again and we're trying to help you and then at 4.45 we're bringing the data to a close with just some quick demos of stuff and any outstanding questions so chlorophyll maps are not without problems because they tend to apply the population's distribution uniformly across the extent of the polygon the census zone and you can see here for example so we've got a bunch of polygons and then what i've done is i've overlaid on top of our piece of aerial photograph and you can see that the actual within the actual real world you have areas of residential here and then over here you have areas of parkland and sport facilities where there's no actual people whereas if you were creating a chlorophyll map and shaded it all red then that would imply that there's people right across the entire zone it's not actually reflecting reality so there's been a trend in recent years for people when creating sort of type maps from census data is to create using alternative approaches called data metric mapping which is this idea where you have the polygons and then you have like a mask layer which contains residential areas and then what you do is you mask the polygons by the mask layer and you create a different form of mapping and one of the websites that does this is I think of datashine which is created by a bunch of folk down in UCL and this is alternative way of viewing census data and you can go to the datashine website and you can browse through all the various census variables and view them using so let me just show you datashine so you've not seen before this is what datashine looks like and you can see what they've done is they've taken all the boundaries and then clip that data based on buildings so you get a much more better representation of where the people live and stuff and it's a really nice website you can just browse through all the various types of census variable and see the various data that's available so you've got a travel for work myth for example and in London you could see travel work by bicycle so that's a nice thing to look at there's also a because the census is delivered separately in England it is from Scotland the data slide different so the datashine folk have a separate datashine Scotland website which uses the same approach but applied to Scottish data we can share those links afterwards if you've not seen this stuff before just go to this and also there's a really nice thing that ONS have produced called their effective maps guide let me just try and find it they published this back in 2018 and it's like a seven page document and it just walks you through I think these are designed for publishing officials to census like statistics but it's got some really nice great practice best practice rubber and it's quite digestible you just walk through it so it tells you about core lift maps it tells you about sort of how many categories you could choose and what's quite classification efforts to use and stuff it's got about well dot maps it's only got stuff about cartograms as well and it gives you advice on when cartograms are best to be used and stuff so it's a really nice thing to look at and I say it's only seven pages so it's really digestible we can maybe look at sharing those as well so that's really all I have to say for today