Implementation:Sdv dev SDV BaseSynthesizer Fit

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

Overview

Concrete tool for fitting a single-table synthesizer on real data, provided by the SDV library.

Description

The BaseSynthesizer.fit method is the primary training entry point for all single-table synthesizers. It preprocesses the raw DataFrame via the DataProcessor pipeline, then calls the model-specific _fit method. After fitting, the synthesizer's internal state is updated to reflect that it is ready for sampling.

Usage

Call this method after initializing any single-table synthesizer (GaussianCopulaSynthesizer, CTGANSynthesizer, etc.) with a metadata object. The data must be a pandas DataFrame matching the metadata schema.

Code Reference

Source Location

• Repository: SDV
• File: sdv/single_table/base.py
• Lines: L675-701

Signature

def fit(self, data):
    """Fit this model to the original data.

    Args:
        data (pandas.DataFrame):
            The raw data (before any transformations) to fit the model to.
    """

Import

from sdv.single_table import GaussianCopulaSynthesizer
# fit is called as: synthesizer.fit(data)

I/O Contract

Inputs

Name	Type	Required	Description
data	pd.DataFrame	Yes	Raw data before any transformations

Outputs

Name	Type	Description
(mutates self)	None	Sets self._fitted = True; internally preprocesses and trains model

Usage Examples

from sdv.datasets.demo import download_demo
from sdv.single_table import GaussianCopulaSynthesizer

data, metadata = download_demo(modality='single_table', dataset_name='fake_hotel_guests')
synthesizer = GaussianCopulaSynthesizer(metadata)

# Fit the synthesizer on real data
synthesizer.fit(data)

Related Pages

Implements Principle

• Principle:Sdv_dev_SDV_Single_Table_Model_Fitting

Requires Environment

• Environment:Sdv_dev_SDV_Python_Runtime