Dashboards are very compact ways of communicating data because they are constrained by space. One way in which we can make more of the "real estate" on the dashboard is to use filters.
Tableau has three ways to filter dashboards. Global filters apply to every part of the workbook that uses the same data source. This might be a problem, however. What happens if you want the filter to apply in some cases, but not others?
Local filters are specific to only one region of the dashboard. However, this may make them too restrictive. Tableau 8.1 now has a new filter feature, which allows you to stipulate a selection of worksheets for the filter rather than being specific to a data source. We can apply the filter to all worksheets that use the data source, and to do this, you can choose the All Using This Data Source option. It is also possible to let the filter apply to only the current worksheet, and you would select the Only This Worksheet option for this.
In this recipe, we will look at the new filter advancements of Tableau 8.1. We will work towards changing a chart into a filled map to show the sales amount, filtering by year. We will add in some new dashboard elements and get them to "talk" to one another by the use of filters. We will apply our filter to selected worksheets rather than the previous editions of Tableau, where it was more "all or nothing" in terms of filtering the data visualization.
For the purposes of this recipe, we will want to select only some of the worksheets, so we'll select the Selected Worksheets option.
In this recipe, we will continue to use the workbook we created for previous recipes of this chapter. There is no need to add more data sources.
- Create a new dashboard called KPI Analysis.
- Take the KPI Summary and KPI Sparkline worksheets and put them in the dashboard by dragging-and-dropping them into place.
- Next, add a title at the top of the dashboard, asking the question
How well did the countries perform? - Place a Blank object underneath the two worksheet objects so that the countries are aligned and read left to right. Your dashboard should now look like the following screenshot:
- When we go to the original KPI Sparkline sheet, we will add the SalesTerritoryCountry filter so that we can filter by country. We do this by dragging the SalesTerritoryCountry attribute to the Filters shelf. The following screenshot is an example:
- This option is fulfilled by selecting the Selected Worksheets option as follows:
- For the purposes of this recipe, we will select the KPI Sparkline worksheets. We do this by simply checking the boxes next to these names. The following screenshot shows an example of the dialog box:
- Once you have clicked on the checkboxes next to the worksheets, click on OK. Once the filters are created, you will need to click on them to display the filters. To do this, click on the filter and choose the Show Filter option.
- Now, go to our Golden Record worksheet.
- To change the Golden Record visualization to a map, simply select the Filled Map option from the Show Me pane.
- Remove Population if it is present, and use SalesAmount to illustrate the sales amount value. In the following example, red has been used because it represents the sales amount for the other components of the dashboard:
- Let's filter the filled map to only show the latest year's data by simply taking FullDateAlternateKey and dragging it to the Filters shelf. When you do this, a Filter Field dialog box will be initiated. We will select the Years filter and will restrict the data so that only the data from the latest year is shown, which in this case is the year 2008. You can see an example of this in the following screenshot:
- The next step, as shown in the next screenshot, is to select the year 2008 from the Filter dialog box; this will filter the map. However, we are not applying this filter to the rest of the dashboard. This means that the map will stay static even though we have filters shared across the other components of the dashboard, KPI Summary and KPI Sparkline.
- Once you have selected the year 2008, click on OK, and we are returned to the map.
- Finally, let's put the Golden Record worksheet with the filled map into the KPI Analysis dashboard. It is static data, and we will place it down the bottom so that it acts as an anchor for the rest of the more detailed data.
The Tableau dashboard will now look as follows:
Filtering data is a key part of the Visual Information-Seeking Mantra article by Professor Ben Shneiderman, and users expect to be able to filter and interact with their data. Although the principles will not hold if people only want very detailed row-level data, the mantra is good to keep in your head when you are thinking about designing your dashboard. Dashboards are about actionable overviews rather than the detail about one row. Hence, the summary is an essential part of providing the overview.
Filtering is a good way to promote engagement with your dashboards. In marketing, stickiness refers to anything that encourages readers to stay on your website. In dashboard creation and reporting, stickiness can refer to features that increase the likelihood that users will stay on your dashboard and use it.
We can use filters in order to make our dashboards more flexible in response to user input, which may help to keep the dashboard engaging and interesting for data consumers. A key aspect to dashboarding is that we need to make the most of the space while engaging the user in the key facts of the data. Filters can help us to do that easily in Tableau.
Once again, color is key to conveying the message of the data. In this example, red is used in both the KPI Summary and KPI Sparkline worksheets, and the color is split into three steps in order to simplify the classification of the sales amount. People don't always distinguish fine-grained nuances of color, and using the Stepped Color feature of the Edit Color panel makes the data simpler to understand.
For the purposes of this recipe, we have only selected some of the worksheets, and the Selected Worksheets option serves our purpose.