DATASETS.md

SDGym Datasets

The datasets used for the SDGym benchmarking process are grouped in three families:

• Simulated data generated using Gaussian Mixtures
  • grid
  • gridr
  • ring
• Simulated data generated using Bayesian Networks
  • asia
  • alarm
  • child
  • insurance
• Real world datasets
  • adult
  • census
  • covtype
  • credit
  • intrusion
  • mnist12
  • mnist28
  • news

Further details about how these datasets were generated can be found in the Modeling Tabular data using Conditional GAN paper, and the code used to generate the simulated ones can be found inside the sdgym/utils folder of this repository.

Using the datasets

All the datasets can also be found for download inside the sgdym S3 bucket in the form of an .npz numpy matrix archive and a .json metadata file that contains information about the dataset structure and their columns.

In order to load these datasets in the same format as they will be passed to your synthesizer function you can use the sdgym.load_dataset function passing the name of the dataset to load.

In this example, we will load the adult dataset:

from sdgym import load_dataset

data, categorical_columns, ordinal_columns = load_dataset('adult')

This will return a numpy matrix with the data that will be passed to your synthesizer function, as well as the list of indexes for the categorical and ordinal columns.

How to add your own dataset to SDGym?

Coming soon!