Principle:Sdv dev SDV Column Distribution Visualization

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

Overview

A visual comparison technique that overlays real and synthetic data distributions for individual columns to assess marginal fidelity.

Description

Column distribution visualization creates side-by-side plots comparing real vs. synthetic data for a single column. Numerical and datetime columns use density/distribution plots (distplot), while categorical and boolean columns use bar charts. This provides intuitive visual assessment of how well the synthesizer captures individual column distributions.

Usage

Use after generating synthetic data to visually inspect individual column quality.

Theoretical Basis

Marginal distribution comparison: overlay $P_{r e a l} (X_{i})$ vs $P_{s y n t h e t i c} (X_{i})$ for each column $X_{i}$ .

Related Pages

Implemented By

Implementation:Sdv_dev_SDV_Get_Column_Plot

Page Connections

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

Principle

Implementation

Heuristic

Environment