SyntheticTabularGeneratorBase<T>
Base Classes · AiDotNet.NeuralNetworks.SyntheticData
Abstract base class for synthetic tabular data generators, providing common infrastructure for fitting models on real data and generating synthetic rows.
Implements: ISyntheticTabularGenerator<T>
Known derived types: AIMGenerator<T>, BayesianNetworkSynthGenerator<T>, CopulaSynthGenerator<T>, SMOTENCGenerator<T>
For Beginners
This is the shared foundation that CTGAN, TVAE, and TabDDPM build on. It handles the “bookkeeping” that every generator needs:
- Remembering column descriptions (which columns are numbers vs categories)
- Managing the random number generator (for reproducibility)
- Tracking whether the model has been trained yet
- Computing basic statistics (min, max, mean, std) for each column
Specific generators override FitInternal and GenerateInternal
to implement their unique training and generation algorithms.
How It Works
This base class handles the lifecycle shared by all tabular generators: column metadata management, random number generation, fitted-state tracking, and the public Fit/Generate API that delegates to subclass implementations.
Properties
| Property | Summary |
|---|---|
IReadOnlyList<ColumnMetadata> Columns { get; } |
|
bool IsFitted { get; set; } |
Methods
| Method | Summary |
|---|---|
Matrix<T> Generate(int numSamples, Vector<T> conditionColumn, Vector<T> conditionValue) |
|
Task FitAsync(Matrix<T> data, IReadOnlyList<ColumnMetadata> columns, int? epochs, CancellationToken ct) |
|
void Fit(Matrix<T> data, IReadOnlyList<ColumnMetadata> columns, int? epochs) |