Merge pull request #49318 from weslbo/power-bi-limited-relationships

SE BUG 144575 - sentence clarification

Author: v-ccolin

learn-pr/wwl-data-ai/choose-power-bi-model-framework/5-determine-when-to-develop-composite-model.yml

@@ -4,9 +4,9 @@ title: Determine when to develop a composite model
 metadata:
   title: Determine when to develop a composite model
   description: "Determine when to develop a composite model"
-  ms.date: 01/11/2023
-  author: wwlpublish
-  ms.author: anrudduc
+  ms.date: 02/26/2025
+  author: weslbo
+  ms.author: wedebols
   ms.topic: unit
 azureSandbox: false
 labModal: false

learn-pr/wwl-data-ai/choose-power-bi-model-framework/includes/5-determine-when-to-develop-composite-model.md

@@ -22,8 +22,7 @@ There are several limitations related to composite models.
 - Import (or dual, as described later) storage mode tables still require periodic refresh. Imported data can become out of sync with DirectQuery sourced data, so it’s important to refresh it periodically.
 - When an analytic query must combine imported and DirectQuery data, Power BI must consolidate source group query results, which can impact performance. To help avoid this situation for higher-grain queries, you can add import aggregation tables to your model (or enable automatic aggregations) and set related dimension tables to use dual storage mode. This scenario is described later in this unit.
 - When chaining models (DirectQuery to Power BI datasets), modifications made to upstream models can break downstream models. Be sure to assess the impact of modifications by performing [dataset impact analysis](/power-bi/collaborate-share/service-dataset-impact-analysis) first.
-
-- Relationships between tables from different source groups are known as limited relationships. A model relationship is limited when the Power BI can’t determine a “one” side of a relationship. Limited relationships may result in different evaluations of model queries and calculations. For more information, see [Relationship evaluation](/power-bi/transform-model/desktop-relationships-understand).
+- Relationships between tables from different source groups are known as limited relationships. A model relationship is considered limited when there's no guaranteed "one" side (The "one" side means the column contains unique values; the "many" side means the column can contain duplicate values.) This can occur in two situations. The first is when the relationship uses a many-to-many cardinality type, even if one or both columns contain unique values. The second is when the relationship spans different source groups, which only happens in composite models. Limited relationships may result in different evaluations of model queries and calculations. For more information, see [Relationship evaluation](/power-bi/transform-model/desktop-relationships-understand).
 
 ## Boost DirectQuery model performance with import data