Derived metrics are constructed from statistical periods, atomic metrics, business filters, and statistic granularity. They are created by applying various calculation logics to existing metrics. For instance, the derived metrics Number of store visits in the last 30 days and Number of purchases in the last 30 days can be combined to calculate the Purchase conversion rate in the last 30 days, which is the ratio of purchases to store visits. This topic guides you through the process of adding derived metrics to a logical aggregate table.
Adding derived metrics
Navigate to the Development section in the top menu bar on the Dataphin home page. By default, you will be directed to the Development page.
Use the operation guide shown in the figure below to access the New Derived Metric page.
On the New Derived Metric page, configure the parameters. For parameter descriptions, see Creating Derived Metrics.
Adding derived metrics
From the Development page, follow the operation guide in the figure below to reach the New Derived Metric page.
In the New Derived Metric configuration wizard, perform the following steps:
On the Define Derived Metric page, input the necessary parameters.
Parameter
Description
Metric English Name
Enter the English name of the metric, adhering to the following naming convention:
Allows letters, numbers, or underscores (_).
Must not exceed 64 characters.
Metric Name
Enter the name of the metric, adhering to the following naming convention:
Allows Chinese characters, letters, numbers, underscores (_), or hyphens (-).
Must not exceed 54 characters.
Data Domain
Select the subject area to which the derived metric belongs.
Statistical Period
Choose the statistical period for the derived metric. Should the desired statistical period be unavailable, create it within the common business logic module. For detailed instructions, see Creating Statistical Period.
Atomic Metric
Select the atomic metric that defines the derived metric.
Business Condition
Choose the business filter that defines the derived metric.
Click Next to proceed.
On the Derived Calculation Logic page, define the calculation logic for the derived metric and use the preview feature to validate the results.
On the Complete Metric Information page, finalize the parameter settings.
Parameter
Description
Data Type
Select the data type for the derived metric. Available types depend on the compute engine; consult the interface for specifics.
Measurement Unit
Choose the measurement unit for the derived metric. Options include currency, time, ratio units, ranking, object quantifiers, weight units, and others.
Currency Units: Cents (RMB), Yuan (RMB), and others.
Time Units: Days, Months, and others.
Ratio Units: Decimal, Percentile, and others.
Ranking: Rank.
Object Quantifiers: Household, Transaction, and others.
Weight Units: Ton, Kilogram.
Others: Various other units.
Business Owner
Select the business owner responsible for the derived metric.
Description
Provide a concise description of the derived metric, within 1000 characters.
Once configuration is complete, use the operation guide shown in the figure below to submit the derived metric.
If your project follows a Dev-Prod pattern, you must publish the derived metric to the production environment. For detailed instructions, see Managing Publish Tasks.
What to do next
After adding and submitting derived metrics or derived metrics to the logical aggregate table, you can go to the Operation Center to manage the logical aggregate table. For more information, see Operation Center.