Connect filters to multiple data sources in ArcGIS Experience Builder

Summary

The Filter widget in ArcGIS Experience Builder is a valuable tool to query and display subsets of data across various data sources. In its standard configuration, the Filter widget is designed to connect to one data source and display a subset of it. However, it is possible to connect a Filter widget to multiple different data sources in the same Experience. This way, the action of selecting one filter toggle can affect different datasets in multiple widgets.

This article describes two ways to connect Filter widgets to multiple datasets and the advantages and limitations of each method.

Procedure

Trigger action

The first option is to create a trigger action in the Filter widget.

1. Add a data filtering action to the experience by selecting Add a Trigger from the Action tab.
2. In the Action settings dialog box, select a trigger dataset and an action dataset. The trigger being the original dataset connected to the Filter widget and the action being the additional dataset to filter.
3. The final step is to add a Trigger connection. Select the matching field in each dataset to filter by. This is similar to a join field. The field name doesn't need to consistently match, but the contents do.

Now every filter in the Filter widget is connected to both data sources. The main advantage to this method is that it is only necessary to configure one action, and all of the filters in the Filter widget will react. Additionally, the More Conditions option below the Trigger connection area allows for complex, multi-lined SQL expressions.

However if you have a Filter widget with multiple filters, and it is necessary for each filter to have unique parameters, the group filter widget may be a better option.

Group filters

The group filter allows for an individual filter within the Filter widget to be set to query multiple datasets. To add a group filter:

1. Go to the Filter widget and select the dropdown arrow next to New Filter.
2. Select the data control directly by this filter. Unlike a trigger action, a group filter can be set to work with more than two datasets at one time.

3. Once the data has been added, create a Group SQL Expression.
4. Select multiple data fields and set whether to have them all satisfy the condition by selecting AND, or whether just one of them, by selecting OR.

The SQL expression that can be built with a group filter is limited in complexity compared to the trigger action. A trigger action allows adding multiple different clauses to the SQL expression, while the group filter only allows a single expression that applies to all of the selected variables.

Which should you use?

Within an individual filter widget it is possible to have a mixture of group filters, standard filters, and custom filters.  So the group filter provides a more specific filter that can be included within a larger Filter widget to query multiple datasets. For example, to have one filter in the widget affect dataset A and B and another filter in the widget to affect dataset C and D, this would be the best option to use.

Alternatively, a filter that includes a trigger action allows creating a more complex query that applies to every filter in the widget. For a complex action that affects datasets A and B to occur for all filters in the widget, a trigger action would be the better choice. One additional note on the trigger action is that while each action can only contain a trigger and an action dataset (so only two datasets affected), it is possible to create multiple trigger actions in one filter widget for the filters to affect datasets A, B, C, and D.

Additionally, when opting to use a trigger action, it only works if the filtered results are less than 1,000 records. If you try to filter a subset that would be greater than 1,000 records, the filter does not run. The group filter option, on the other hand, doesn't have a documented limit like this. Users who are trying to filter for large subsets of data, should use the group filter instead.

On the surface, the group filter and trigger action options seem similar in terms of functionality. However, each method provides unique benefits and their differences should be considered. Ultimately, workflows that include complex analysis on smaller data would be ideal for a trigger action. Users who need a variety of filters in specific widgets should use a group filter instead, especially if they are working with large datasets.