A filter is used to provide different combinations of data. Instead of having a separate visual or dashboard for each category, you can add filters and let users change their criteria to explore the data. They allow you to view the same dashboard from a different business viewpoint. Bold BI allows you to configure different types of filters in dashboards, as well as preserve your filtered data to explore at any time. You can add filters at the data-source level, restricting the data that will be displayed to each user. You can also showcase top- or bottom-performing records by using data-source-level filters or widget-level filters in Bold BI. In this blog post, we will see different filters supported in Bold BI and their uses, and how to configure them to create multiple views of data using a single dashboard.
Now, let’s examine each supported filter and its use case in Bold BI dashboards.
Supported filters in Bold BI dashboards
A data-source-level filter allows you to add a filter in the data source designer configuration that restricts the visibility of records based on defined criteria before you have designed the dashboard. It allows you to apply filtering to the data types, including Text, Numeric, DateTime, and Boolean type columns.
In the Northwind Products and Suppliers Dashboard, you can monitor the sales performance for available products alone by applying a filter to the data source. The discontinued product record is restricted by selecting False for the Discontinued column in the Query Filters window as shown in the following image.
Only current product records are showcased in the data preview grid as shown below.
Next, in the same dashboard, I am going to add another filter condition in the data-source filter to showcase only the current product. The following dashboard screenshot shows the sales performance of all products. Note that the overall sales amount shown in the number card widget is $753,306.
Here, you can restrict the dashboard visuals by showing only the current product order summaries by sales team. The following dashboard screenshot shows sales performance for the current product only.
You can see the total sales is $623,829, which is the available product’s sales amount. To achieve this view, you can add another or multiple filters in the Query Filters configuration. For the text data type column, the parameter shows the specific conditions like Starts With, Ends With, etc. In this example, the specific condition is Inclusion, as shown in the following image. Similarly, for other data types, the relevant condition will be displayed to create query filters.
For the integer type column, the parameter will be as shown in the following image.
Check out our data filters documentation for more details on configuring filters for your dashboards.
User-based filtering is the process of imposing row-level security on the underlying data, thereby providing user-based data access. User-based filters help you avoid re-creating the same dashboard for each user; you can maintain one dashboard for all users by restricting other users from the information. This blog provides a brief discussion about user-based filtering.
The Motor Vehicle Crashes Analysis Dashboard shows the number of accidents in all states in a map widget. As an example, let’s say that management decides to provide access to the dashboard for each regional manager so they can visualize their region’s data and make changes to make driving safer.
The following dashboard screenshot shows the administrator view, which shows the crash data for all regions.
The following screenshot shows a regional manager’s view, which shows only their own state’s data.
You can see that Indiana is the highlighted state in the map widget for this user, and all the other widgets only display the data for Indiana. Refer to the user-based filtering documentation for more details about this type of filtering in Bold BI.
A widget-level filter is applied to data as measures and dimensions. Measure fields comprise numeric types whereas dimension fields comprise strings, dates, and other types.
In the Online Marketing Dashboard, we showcase website activities and other key metrics. By selecting a dimension filter in the Channel Name column, we can apply item-based, condition-based, and rank-based filtering. These allow us to track things like the top five revenue traffic sources, specific revenue ranges, and a specific list of traffic sources.
You can track revenue from a specific list of traffic sources by using include or exclude conditions.
Also, you can apply filters based on conditions to track specific revenue ranges.
You can filter the top five values by applying the Rank filter to the Revenue column in the Filters window as shown below.
You can see the top five revenue traffic sources in the pie chart, which is showing the top five traffic sources based on revenue.
Let’s look at an example of measure filters. The Insurance Analysis Dashboard showcases the performance of an insurance company. By selecting a filter from the measure field for the Profit column, you can prevent negative values from being showcased in the widget. Set the filter Greater Than 0 to show positive values only, as shown in the following image.
The following dashboard screenshot shows the performance of the whole insurance company. Note that the Profit vs. APE bar chart has negative values.
After setting the filter to ignore negative values, the following screenshot shows the Profit vs. APE chart without negative values.
Next, let’s see how you can track the performance of a company over any relative dates. For example, if you want to filter the last 30 days, last week, or any specific date ranges, you can do so by applying relative filters to the Date column.
The following Retail Store Performance Dashboard screenshot shows the performance of apparel stores.
Note the sales date in the date picker is 9/20/2020 to 9/25/2020
The following screenshot shows the retail stores’ performance for the last 30 days, which is a relative date range.
You can see the sales date is 8/26/2020 to 9/24/2020, which is the last 30 days. You can check out more widget filter configurations in our documentation.
Now let’s explore the dynamic filter options available in widget-level filtering. They are:
Allow filtering enables filtering within a widget. Tabular-format widgets allow you to filter data in each column of the widget. You can perform a dynamic filtering operation on data within a widget by enabling the Allow filter option in the properties for the grid and pivot grid widgets.
For the grid, a filter box will be enabled in each column as shown below.
For the pivot grid, a filter icon will be enabled as shown.
The Pharmaceutical Production Analysis Dashboard showcases KPIs relevant to the drug manufacturing process. It tracks overall production quality and the number of drugs produced compared to the production target, among other things. When tracking drugs that have many unique categories in a column, you can focus on specific categories through the allow filtering option.
The following dashboard image shows overall production and the production targets.
Note that the drugs shown in pivot grid is an overview of all the drugs. In the image below, the pivot grid data is filtered to show only the drug production and its target value for Syrup 1.
A hierarchical filter helps when applying a top N filter with multiple dimension columns. The filtering is performed based on the hierarchy of the dimension columns. You can perform filtering based on hierarchy level in the grid and pivot grid widgets only.
The following Student Details Dashboard visualizes various aspects of a student’s performance that is helpful for teachers and administrators.
In the following screenshot, I have applied a top N filter to the student column to get the names of the top two students.
By enabling a hierarchical filter, the top two students are filtered from each branch as shown below.
Without hierarchical filtering, the top two students are filtered and those students’ details across all branches are shown, as in the following image. The students Bart Gusman and Emerson Butler are filtered.
Drill-down filtering is multilevel filtering that you can use to view detailed breakdowns of your data with just a click. In the Personal Expense Analysis dashboard, a drill-down filter is applied to the doughnut chart shown in the following screenshot. The Home category is highlighted because we can click on it to drill down into its data.
After navigating into the multilevel breakdown of the Home category, you can see a more detailed view of home expenses as in the screenshot below.
You can refer to the documentation to learn how to enable drilling down in widgets.
Dashboard filters help apply master and listener relationships between widgets. They help control the interdependency of widgets in a dashboard with respect to dynamic user interactions.
Enable the act as master property in a widget you want to act as a master. Select the listener widgets that you want to participate in filtering in the Filter Configuration window as shown below.
The following Sprint Management dashboard showcases key metrics for tracking sprint performance.
Note the project selection combo box, which is currently set to all projects.
The following screenshot shows sprint performance for Project A alone.
Managers can track sprint performance for individual projects by applying a master filter.
Bold BI allows you to use widgets in filtering from one or more data sources. You can achieve this by enabling the Custom option in the Filter Configuration window. You can map a column from the current data source to the target data source as shown below.
The following GitHub dashboard shows GitHub metrics that track activities, pulls, and commits. These activities are fetched from an individual data source. However, those activities can be tracked for all repositories through mapping the columns to a target data source.
The following screenshot shows the activities of one repository.
Refer to this dashboard filters documentation for more details about configuration in Bold BI.
Finally, Bold BI allows you to save the filtered view after publishing the dashboard and preserve your filtered data to view it at any time. For example, if you are filtering your dashboard view using a combo box or any other filter widgets, the resultant view will be saved in the Filter Overview option available in the top right corner of the page.
If you want to preserve a filter applied to a dashboard, you can save the view through the Save option under Filters Overview.
We hope this blog article provided valuable information about the various data filtering options in dashboards to help you deliver a seamless user experience. If you have any questions about data filtering, please post them as comments on this blog. You can also contact us by submitting your questions through the Bold BI website or, if you already have an account, you can log in.