Features

Synthia includes several software features for the generation and augmentation of multivariate data (Table 1). You can find out more about how to use these option in the examples page.

Table 1. Software features included in the latest version of Synthia for different fitting and generation methods. For correlated multivariate variables stretching and uniformization is available for copula models (^). All models support the handling of continuous variables (¹). Discrete (²) and categorical (³) variables are supported with vine copulas. All correlated multivariate methods support empirical and parametric fitting of marginal distributions

Variable relation	Method	Option
Univariate or Multivariate Independent
	Fitting
		Empirical
		Parameterized
	Generation
		Pseudo-random
		Quasi-random
		Stretching
		Uniformization
Multivariate
	Fitting
		^¹Copula: Gaussian
		^¹²³Copula: Vine
		¹fPCA
		Empirical
		Parameterized
	Generation
		Pseudo-random
		Quasi-random
		^Stretching
		^Uniformization