[MUSIC] Welcome back. In the last lesson, we reviewed our data set and came up with the questions that we're going to answer based on our quick view of the data set and how our made-up context where we're executives of a major London public transportation office. In this lesson, you'll be able to create a bar chart which will help us find patterns in our data set. So let's get going and remember that if you need to stop at any point, make sure you save your work to Tableau public so it'll be easy to pick up where you left off. We're going to tackle all of the questions from lesson one, so we finally get to start digging in. Let's open up Tableau and the workbook we saved a Tableau public last week, then go to Sheet 1 to get started. Our first question is about location. Now, we don't actually have any real mappable location information like latitude and longitude in this data. But we do have columns for borough, bus garage, group name, and operator and the actual route number. We could go crazy and track them all the latitudes and longitudes for the routes and boroughs, but let's not get ahead of ourselves here, there's still plenty we can learn from what we have. We want to know the number of injuries, so we'll use the Number of Records field, which we already conveniently added to the view last week. Boroughs seems like a good place to start for a where so let's drag that out onto the rows, and we'll get a nice list of London boroughs sorted alphabetically because we haven't told it any other kind of sort yet. We can see that Tableau automatically did a group by at the borough level, and it's displaying the number of incidents for each borough. We can kind of eyeball where the most injuries might be, but we still have to look pretty hard and read every one of those four digit numbers to see which is the largest. Not a very efficient way to use our precious executive time, clearly this is not the best way to display this data, but probably a bar chart would be good. Let's take that green pillow on the marks card, which says, sum number of records, and drag it up to the columns. Doing this is telling Tableau that we want the columns, or the information going across the view from left to right to indicate the number of records, Tableau than automatically makes a bar chart for us. If you look back at the marks card, you could see that we didn't do anything to tell Tableau that we wanted bars, but it took a good guess and thought that would work pretty well, and it does. Try changing the drop-down menu to some other chart type, and you'll see that the bars really are the best for the job. Okay, we're making really good progress here but this bar chart looks really messy, doesn't it? Sure it's pretty easy to scan and find the biggest bar but what's the next biggest and then the third biggest, hard to tell without looking at all of them. So like we talked about earlier let's sort them in descending order. This quickly shows us that Westminster had the most bus related injuries during the time this data was collected and by a decent amount too. Let's go ahead and hide this show me menu because we don't really need it right now. Up next almost exactly tied are Southwark and Lambeth, great we've answered our first question. Let's rename this sheet, Incidents by Borough. Then right click, duplicate, and go to the new sheet you just made. The next piece of this question is about routes, use this new sheet to build a view about the number of incidents by route. I would hazard a guess that OOS means, out of service, and we look at the underlying data for this that's probably the case. There's nothing that stands out saying that this is actually a route because it is present in many boroughs and with many different operators. That's okay though, it's still data and they're still injuries, so let's leave it there. Finally, we wanted to know about types of incidents, let's rename this sheet as Incidents by Route. And then go back to the first sheet and duplicate it again. We wanted to know about locations of various incident types, so again Borough seems like a good place to start. So far we've had just one pillow on the rows shelf, but what if we add another? Click and drag, Incident Event Type, to the right of the borough pill on the rows shelf? Now we have the number of incidents grouped by borough and then again by incident event type. Let's do our usual sorting and see that Tableau sorts the events independently within each Borough. Count yourselves as lucky because this kind of nested sorting only came out with version 2019.1. Back in my day we didn't have nested sorting [SOUND] but in all seriousness now, we can easily see the most prevalent incident type in each Borough. Let's rename the sheet, Incidents by Borough and Type, and keep going to the next set of questions. Remember to save your work to Tableau public if you need to stop here.