This repository contains a collection of datasets for causal discovery, which are split into two folders: sparse and dense. Each folder contains datasets that exhibit either sparse or dense causal DAGs.
The sparse folder contains datasets that have a small number of causal relationships between variables. These datasets are further split into two subfolders based on the type of causal mechanism: mixed causal mechanism and linear causal mechanism.
The dense folder contains datasets that have a large number of causal relationships between variables. These datasets are also split into two subfolders based on the type of causal mechanism: mixed causal mechanism and linear causal mechanism.
This subfolder contains datasets that exhibit a mix of causal mechanisms, including both linear and nonlinear relationships. The datasets are split by different types of issues and dataset size.
This subfolder contains datasets that exhibit linear causal relationships between variables. The datasets are split by different types of issues and dataset size.