KB483883: Best practices for contextual linking in dossiers

Starting in Strategy 2020, you can use a visualization in one dossier to filter data in a different dossier. First, define one visualization as the source. Then, select a page from a different dossier that you want to filter as the target. The target visualizations only display data that also appears in the source.
When you choose to filter data within another dossier, you also have the option to pass the filter conditions in the source dossier to the target dossier. However, to successfully pass the filter, there are some best practices you should follow.

Rules for passing filter conditions in the Filter panel

When you choose to filter data within another dossier, you have the option to pass the filter(s) under the Filter panel in the source dossier to the target dossier. The following matching rules apply for attribute and metric filters:

Attribute Filters

1. Both the attribute object ID and the filter name in the source and target dossiers must be the same.
   The filter name refers to the attribute name that is displayed inside the Filter panel. Generally, this corresponds to the attribute itself. However, it is possible to rename an attribute via the Datasets panel (right-click > Rename) and then add it to the Filter panel. In this situation, the new name becomes the filter name, potentially causing a mismatch between the two dossiers.
   For example, if the same attribute, Date, exists in the source and target dossiers and has been placed in the Filter panel, the filter name is consistent (Date). However, if you rename Date to Date1 via the Datasets panel in the source dossier and then remove the attribute from the Filter panel and add it back, the new filter name in the source dossier is Date1. If the filter name is different in the source and target, the filter condition is not passed.
   
2. For scenarios where the attribute object ID is not the same, the filter name must be the same.
   This is a common scenario when the source and target dossiers use different datasets created via Data Import, where the imported data may be the same, but the attribute object IDs between the two datasets is not the same. In this case, filter selections can still be passed between a source and a target dossier if the ID form of the attributes share the same type (e.g. number, date-time), and if the filter names match as explained above. For example, if an attribute named Category is imported from two external datasets, such as Excel files, and both contain the same ID form and data type for Category, such as number, the  Filter panel selections are passed successfully.
   

Metric Filters

Similarly, if the metric filter name and the metric object ID are the same, the filter condition can be passed from the source to the target dossier.
Dossier linking also supports passing selections made in a visualization in the Filter panel. For this situation, similar rules apply to as those mentioned in the Attribute Filters section above. The filter name should be the same between the source and target.

Rules for passing data selections made in a visualization to another dossier

When you select a data point from a grid or visualization in the source dossier, the selected data point(s) are automatically passed from the source to the target dossier as a view filter to the visualizations contained in the page of the target dossier. The following rules apply for passing data selections as filters:

1. If the attribute ID in the source and target dossier are the same, selections are passed.
2. If the attribute ID in the source and target dossier are different, then the attribute name must be the same for the selections to be passed.

Note that for rule 2, there are three levels in which an attribute names can be defined in the product: at the individual attribute object level (applicable to project schema attributes), at the report or cube level, and at the level of the dossier (via the Datasets panel).

• For example, the attribute name (original object name) is called, Category. Rename it to Category_alias in dataset in the report: 
  

• Rename it to Category_alias in dossier in the Datasets panel of the dossier:
  

• The following table compares the different rule level attribute names from a source to target dossier.
  Where a is the attribute name in the Datasets panel of a dossier.
  Where b is the attribute name in the original dataset (report/cube).
  Where c is the attribute name (original object name).
  

• The least optimal condition would mean the selection is passed and compared nine times until c in the source is the same as c in the target.
  If any of the conditions in the table above are matched, such as a –> b or b –> c, then compare the attribute's ID form, including the number and type. If they are all the same, the selection can be passed from the source to the target dossier.
  

For more information on how to use visualizations to filter data in a different dossier, see the MicroStrategy Workstation Help.