Derived metrics summarize a business activity over a fixed period, but they cannot express relationships between periods. For example, you cannot use a derived metric alone to calculate week-over-week growth or a year-over-year change ratio. Composite metrics are built on top of derived metrics using configurable computation rules, letting you define cross-period calculations such as growth rates, period comparisons, and weighted ratios.
Prerequisites
Before you begin, make sure you have:
-
Created the derived metrics that the composite metric will reference. See Create a derived metric.
-
Created a data layer to store the composite metric. Choose the layer based on your requirements. See Define data warehouse layers.
-
Data Warehouse Summary (DWS) layer: for composite metrics in the common layer
-
Application data layer: for composite metrics in the application layer
-
-
Created a business process, data mart, or subject area to categorize the metric:
-
Composite metric (common layer) requires a business process. See Business processes.
-
Composite metric (application layer) requires a data mart or subject area. See Data marts or Subject areas.
-
Go to the Composite Metric page
-
Log on to the DataWorks console. In the top navigation bar, select the desired region.
-
In the left-side navigation pane, choose Data Development and O&M > Data Modeling. Select the desired workspace from the drop-down list and click Go to Data Modeling.
-
On the Data Modeling page, click Data Metric in the top navigation bar.
Create a composite metric
In the upper-right corner of the Composite Metric list, click
to open the creation panel.
Choose a computing logic
Select one of the following computing logic methods based on your use case:
|
Method |
When to use |
How it works |
|
Expression |
Flexible calculations — growth rates, ratios, weighted averages, or any custom formula |
Write a calculation expression. Use |
|
Comparison |
Year-over-year or period-over-period changes only |
Select a derived metric and a comparison period. The system computes the change automatically. |
Set basic information
Select whether the composite metric belongs to the common layer or the application layer, then follow the on-screen instructions to set the business scope. The available parameters vary by layer.
Save and commit
After completing the configuration:
-
Click Save to save the current version.
-
Click Commit to publish the version.
Only saved composite metrics can be committed. Only committed composite metrics can be referenced by other metrics. Each commit creates a new version record; a committed version cannot be committed again.
Manage versions
Open the right-side navigation pane on the metric configuration page to view the version history and perform the following operations.

Version comparison
Compare the content of two versions to identify what changed:
-
Select one version to compare it against the currently saved metric content.
-
Select two versions to compare them directly. Two versions is the maximum.
Rollback
Switch the metric configuration page to a previous version. The current content on the page is overwritten by the selected version.
Rollback updates only the UI configuration. The underlying data models and derived metric references are not updated until you save and commit again after rolling back.
Batch operations
Click
in the upper-right corner of the Composite Metric tree to switch to list view. Select multiple metrics to perform the following batch operations.
Batch export
Export composite metrics using one of three options:
-
Export All: Exports all composite metrics in the current workspace.
-
Export Searched Objects: Enter a keyword to filter metrics by name (fuzzy search is supported), then export all matching results.
-
Export Selected Objects: Exports the metrics you have selected.
After the export completes, download the exported file to your local computer.
Batch delete or submit
-
Batch Delete: Removes selected metrics. Deleted metrics can no longer be associated with or referenced by model tables. If a metric is currently referenced by a model table, remove the reference relationship before deleting.
-
Batch Submit: Commits selected metrics so they can be associated with and referenced by model tables. A new version record is generated for each commit. For details, see Manage versions.