Principle:Sdv dev SDV Cardinality Analysis

Knowledge Sources	SDV Documentation SDV
Domains	Visualization, Relational_Data, Data_Quality
Last Updated	2026-02-14 00:00 GMT

Overview

A visualization technique that compares the distribution of parent-child row counts between real and synthetic relational data.

Description

Cardinality analysis visualizes how many child rows each parent row has, comparing real vs synthetic data. This reveals whether the synthesizer correctly reproduces the distribution of relationship cardinalities (e.g., how many orders per customer, how many items per order). It is specific to multi-table data.

Usage

Use after generating multi-table synthetic data to verify that parent-child relationship cardinalities are preserved.

Theoretical Basis

For a parent-child relationship with foreign key FK:

Compute cardinality per parent: $C_{i} = | r o w s \in c h i l d : F K = P K_{i} |$
Compare distributions of $C_{r e a l}$ vs $C_{s y n t h e t i c}$

Related Pages

Implemented By

Implementation:Sdv_dev_SDV_Get_Cardinality_Plot

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment