Keysets are collections of unique data values that can be saved and used in further analysis of your data. They allow you to perform multipass and multisource filtering of your data.
You can create a keyset by selecting a single field for the keyset from visual data. Filtering the visual data before creating the keyset limits the unique data values the keyset includes. The field you select for the keyset is known as the keyset's key field. For example, from a visual that lists airport codes nationwide but that has been filtered to show only Virginia airports, you can create a keyset listing only Virginia airport codes by selecting the airport code field as the keyset's key field.
You can apply a keyset as a visual filter. This limits the visual's result set to records containing values stored in the keyset. Keyset filters can be applied to visuals from the same or different data sources. However, the data source for the target visual must contain a field storing the same kind of data as the keyset's key field.
When working with multiple filter types on the same visual, row-level filters (including wild card filters) are applied first to a visual, keyset filters are applied second, and group filters are applied last on the aggregated result set.
If you change the field type from Number or Integer to Attribute for a field that is used in a keyset, errors will result when the keyset is applied as a filter. To resolve this problem, create an Attribute derived field from the original field (for example, using the
num_to_text function) and use it when you create the keyset, instead of the original field.
To control whether keysets can be created or managed from a visual, use the interactivity sidebar. See Controlling How Users Interact With a Visual.
See the following topics.