Logo

Getting Started

  • Overview
  • Features
  • Installation
    • Required dependencies
    • Optional dependencies
    • Instructions
  • Quickstart
    • Load required modules
    • Define plotting function
    • Load and plot sample data
    • Fit and generate 1000 random samples using Gaussian copula
    • Fit and generate 1000 random samples using fPCA
  • Examples
    • Generation
      • Univariate
        • Import libraries
        • Load sample data
        • Fit and generate using samples
        • Fit and generate using parametrized distribution
        • Stretching and unifomization
        • Exporting and saving generated data
        • Saving the generator
      • Multivariate: Independent
        • Import libraries
        • Create test data
        • Generate synthetic data
        • Generate synthetic data with modified characteristics
        • Parameterize data using quantiles
        • Parameterize data using distributions
      • Multivariate: Gaussian Copulas
        • Import libraries
        • Create a sample dataset with n samples
        • Fit a Gaussian copula with Synthia’s backend
      • Multivariate: Vine Copulas
        • Import libraries
        • Create a sample dataset with n samples
        • Fit a Vine copula with pyvinecopulib’s backend
      • Multivariate: fPCA
        • Import libraries
        • Define plotting function
        • Plot source data
        • Fit the fPCA model using 10 components
        • Generate same number of samples as in the input
        • Plot the results
      • Multivariate: Discrete and Categorical
        • Import libraries
        • Generate dummy data
        • Fit and generate new samples
    • Enhancement
      • Stretching and Uniformization
        • Import libraries
        • Define plotting function
        • Plot source data
        • Fit copula to data
        • Generate ‘streatched’ samples
        • Generate ‘more uniformly distributed’ samples

Background

  • Copulas
    • What copulas are
    • The Gaussian copula
    • Other copula families
    • Vine copulas
  • Functional Principal Component Analysis (fPCA)
    • The general idea
    • Mathematical definition
    • PCA as a basis expansion
    • PCA for synthetic data generation

Help & reference

  • API reference
    • Data generators
    • Copulas
    • Parameterizers
    • Transformers
    • Utilities
  • How to cite
  • Contributing
  • Development notes
    • Conda environment
    • Install synthia
    • Documentation
    • Docstrings
    • Testing
    • Versioning
    • Deployment
  • Copyright & License
synthia
  • Docs »
  • Search
  • Edit on GitHub

© Copyright 2020 D. Meyer and T. Nagler

Built with Sphinx using a theme provided by Read the Docs.