Data standardization and modeling

Dataphin follows Ralph Kimball's dimensional modeling theory. You design conceptual models based on your business, then create dimension tables, fact tables, atomic metrics, business filters, metrics, and logical aggregate tables from the model's business entities (business objects and business activities).

Prerequisites

The project is bound to a data section. Projects with incorrect or missing bindings cannot use data modeling. Bind a project in Create a General Project.
The default feature menu items selected during project creation include data modeling modules. Configure them in Create a General Project.

Note
Basic projects linked to Basic Mode data sections support data modeling. Basic projects linked to Dev-Prod Mode data sections only support data processing and ad hoc queries.

Features

The following table describes each modeling feature:

Feature item	Description	References
Logical dimension table	Logical dimension tables store attribute details of business objects. Create tables matching each business object type.	Create a General Logical Dimension Table. Create a hierarchy dimension logical table. Create a logical dimension table for enumeration. Create a virtual dimension logical table.
Logical fact table	Logical fact tables store data generated by business activities. Create tables matching each business activity type.	Create a logical fact table.
Atomic metric	Atomic metrics define the statistical scope and calculation logic for a business measure. Standardized metric definitions ensure consistent results and improve R&D efficiency. Example: payment amount.	Create atomic metrics.
Composite metric	Composite metrics derive calculations from multiple atomic metrics. Example: C=A/B.	Create a composite metric.
Derived metric	Combines an atomic metric with business filters, a statistical period, and dimensions to produce a business statistic. Derived metric = Atomic metric + Business filter + Statistical period + Dimension(s) (statistic granularity).	Create a derived metric.
Business filter	Business filters filter records by business rules, similar to SQL WHERE conditions (excluding time intervals). While atomic metrics define calculation logic, business filters define condition constraints.	Create business filters.
Logical aggregate table	Logical aggregate tables group derived metrics that share the same statistic granularity and period. Each derived metric is associated with a single logical aggregate table. The primary key consists of dimensions from logical dimension tables (the statistic granularity); all non-key fields are metrics.	Create logical aggregate tables. Configure a logical table conversion task.

Access data modeling

On the Dataphin home page, click R&D in the top menu bar.
The left navigation pane of the Data Development page lists entry points for each data modeling module.