Iterations of bar charts for deeper analysis

Using the preceding bar chart, you can easily see that the Technology department has more total sales than either Furniture or Office Supplies, which has fewer total sales compared to any other department. What if you want to further understand sales amounts for departments across various regions?

  1. Navigate to the Bar Chart (two levels) sheet where you will find an initial view identical to the one you created previously.
  1. Drag the Region field from Dimensions in the data pane to the Rows shelf and drop it to the left of the Department field already in the view, as shown:

You still have a horizontal bar chart. But now you've introduced Region as another dimension that changes the level of detail in the view and further slices the aggregate of the sum of Sales. By placing Region before Department, you will be able to easily compare sales for each department within a given region.

Now you are starting to make some discoveries. For example, the Technology department has the most sales in every region, except in the East where Furniture has higher sales. Office Supplies never has the highest sales in any region.

Let's take a look at a different view, using the same fields arranged differently:

  1. Navigate to the Bar Chart (stacked) sheet where you will find an initial view identical to the one you created previously.
  2. Drag the Region field from the Rows shelf and drop it on the Color shelf:

Instead of a side-by-side bar chart, you now have a stacked bar chart. Notice how each segment of the bar is color-coded by the Region field. Additionally, a color legend has been added to the workspace. You haven't changed the level of detail in the view, so sales is still summed for every combination of region and department.

The Level of Detail or View Level of Detail is a key concept when working with Tableau. In the most basic visualizations, the combination of values of all the dimensions in the view defines the lowest level of detail for that view. All measures will be aggregated or sliced by the lowest level of detail. In the case of most basic views, the number of marks (indicated in the lower-left corner of the status bar) corresponds to the number of intersections of dimensional values. If Department is the only field used as a dimension, you will have a view at the department level of detail and all measures in the view will be aggregated as per the department. If Region is the only field used as a dimension, you will have a view at the region level of detail and all measures in the view will be aggregated as per the region. If you use both Department and Region as dimensions in the view, you will have a view at the level of department and region. All measures will be aggregated per the unique combination of department and region.

Stacked bars are useful when you want to understand part-to-whole relationships. It is now fairly easy to see what portion of the total sales of each department is made in each region. However, it is very difficult to compare sales for most of the regions across departments. For example, can you easily tell which department had the highest sales in the East region? It is difficult because, with the exception of West, every segment of the bar has a different starting place.

Now, take some time to experiment with the bar chart to see what variations you can create:

  1. Navigate to the Bar Chart (experimentation) sheet.
  2. Try dragging the Region field from Color to the other shelves on the Marks card, such as Size, Label, and Detail. Observe that in each case, the bars remain stacked but are redrawn based on the visual encoding defined by the Region field.
  3. Use the Swap button on the toolbar to swap fields on Rows and Columns. This allows you to very easily change from a horizontal bar chart to a vertical bar chart (and vice versa):
  1. Drag and drop Sales from the Measures section of the data pane on top of the Region field on the Marks card to replace it. Drag the Sales field to Color if necessary and notice how the color legend is a gradient for the continuous field.
  2. Further experiment by dragging and dropping other fields onto various shelves. Note the behavior of Tableau for each action you take.
  3. From the File menu, select Save.
At the time of writing of this book, Tableau does not have an autosave feature. You will want to get in the habit of saving the workbook early and then pressing Ctrl + S or selecting Save from the File menu often to avoid losing your work.
..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset