From the course: Complete Guide to Tableau for Data Scientists

What is the Analytics pane? - Tableau Tutorial

From the course: Complete Guide to Tableau for Data Scientists

Start my 1-month free trial Buy for my team

What is the Analytics pane?

“

- [Instructor] Tableau not only makes it really easy to visualize our data, it's increasingly making it easier to analyze it as well, giving us drag and drop analytics when we need it. Now, this is very simple to do, because this is how Tableau tries to everything to make it simple as possible. But what is the analytics tab and what do we do with it? Well, if we build a simple chart, we use just drag and drop. So we drag the fields that we are interested in to build a chart. I put the sales on one axis, I put the category on another, and I get a bar chart, because that's the rules that Tableau has. But if I want to go one step further, I might need to do some further analysis on my visualization after I've created it. And this is where the Analytics tab comes in. Now we can find it at the top of the data pane, and in fact, you probably didn't even notice it was there. If we click on Analytics, it changes to the analytics pane on the left hand side. These are secondary analytics that we apply once we've built our view. Now that's really important to understand, and it's actually why it's on a separate tab. Previously, anything that we've done within the data tab, we've added or modified a column in our data. We've either changed the values of it. We've maybe set the default values. We might have created the calculation which creates an extra field, which in a sense, creates an extra column within our data set. Now, all the time we've done that, we've done that within the data pane. This time we're looking at the analytics pane. So this gives you some understanding of where the analytics happens. The data pane contains all of our raw data, plus all of our modified data that we then drag and drop into our view. Now, the combination of the rows and the columns, the measures, dimensions, discrete, continuous builds the kind of view. We can control what the marks look like with the marks card. Now, once we've got that view built, we can then drag and drop the analytics onto it. So that's why they're separated out, to show that we build the views and the data, then we analyze the view using the analytics. So the analytics part comes at the end of that process. Once we've created our view, we can then start adding the analytics to it. For example, we might want to say, what is the average sales across all of our data? Now, I can't do that for my original data. I could create a calculation, but I want to see it visually. So what Tableau allows us to do is drag and drop things like an average line. I could take the average line out of my analytics pane and I could drop it over the entire table of data. Tableau then draws a line against the sales axis to tell me that the average sales for all of my data is 136,000. Now, that value is dependent on what data I've included as part of my view. If I filtered out some of that data, that average is going to take that into account. Depending on what measures and dimensions I have in the view, this could change the kind of analytical answers that I get. This is because anything that we add from the analytics tab happens after the view has been created. In a way, they are versions of, or very similar to table calculations that we've looked at in the calculations chapter. These happen after the view has been built. So after everything else has been taken into account, once we have an axis, we can then add the reference line. Now, this is because reference lines are attached to the axis. But we can also add things such as totals and subtotals. That could happen with either a bar chart such as this, or maybe with a table. Once we've created our view, we can also do things like modeling and forecasting. This allows us to look at our data, but only after we've visualized it, because we need something to look at. We can't create a trend line or a forecast for this particular data type, because it just doesn't make sense. You can't create a forecast based on a bar chart. The forecast and the trend line need a timeline to work with. So we can only add some of these analytics when we have the correct view. This is another indication to show that the analytics happens after we've created our initial visualization. For more detail on how to create each one of these analytics, check out the further videos in this chapter. Remember, we have to create a visualization first, then we can drag and drop the analytics on top of it to further understand and see what's happening with our data.

Contents