Implementation:Sdv dev SDV HMASynthesizer Init

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

Overview

Concrete tool for creating an HMA-based multi-table synthesizer that preserves relational structure, provided by the SDV library.

Description

The HMASynthesizer implements the Hierarchical Modeling Algorithm. It creates per-table GaussianCopulaSynthesizer instances, manages table augmentation with child statistics, and coordinates hierarchical sampling.

Usage

Import this class when you need to generate synthetic multi-table relational data. It requires a Metadata object with defined relationships.

Code Reference

Source Location

Repository: SDV
File: sdv/multi_table/hma.py
Lines: L187-206

Signature

class HMASynthesizer(HMASynthesizerBase, BaseHierarchicalSampler):
    def __init__(self, metadata, locales=['en_US'], verbose=True):
        """
        Args:
            metadata (Metadata): Multi-table metadata with relationships.
            locales (list): Locale(s) for AnonymizedFaker. Defaults to ['en_US'].
            verbose (bool): Print progress during fit/sample. Defaults to True.
        """

Import

from sdv.multi_table import HMASynthesizer

I/O Contract

Inputs

Name	Type	Required	Description
metadata	Metadata	Yes	Multi-table metadata with relationships defined
locales	list	No	Locale(s) for fake data (default: ['en_US'])
verbose	bool	No	Print progress (default: True)

Outputs

Name	Type	Description
instance	HMASynthesizer	Unfitted multi-table synthesizer

Usage Examples

from sdv.datasets.demo import download_demo
from sdv.multi_table import HMASynthesizer

data, metadata = download_demo(modality='multi_table', dataset_name='fake_hotels')

synthesizer = HMASynthesizer(metadata)
synthesizer.fit(data)
synthetic_data = synthesizer.sample(scale=1.0)

for table_name, df in synthetic_data.items():
    print(f"{table_name}: {len(df)} rows")

Related Pages

Implements Principle

Principle:Sdv_dev_SDV_HMA_Synthesis

Requires Environment

Environment:Sdv_dev_SDV_Python_Runtime

Uses Heuristic

Page Connections

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

Principle

Implementation

Heuristic

Environment