Scope
We accept datasets and statistical software products that are relevant for the task of structure learning in the context of graphical modelling and causal discovery. This is equivalent to learning an underlying graph from data. A typical example of a dataset would be gene expression data where each is a data vector holding simultaneously measured expression levels for a collection of genes. Structure learning then holds the potential to shed light on learning regulatory relationships between genes.

Format of data
All datasets should be in tabular format: .csv, .txt, .tsv. Additionally, if a ground truth exists for the dataset (i.e., the model describing the true structural relationships between the variables), it should be provided in a separate file. This file should contain a list of edges in a format readable by typical graphical model packages (e.g., bnlearn, iqgraph in R). We welcome submissions of versatile dataset collections which is or can be used in the field of graphical modelling.

Metadata
Metadata are provided and stored in a separate markdown and in html file.

Access
All content is supplied with a license that outlines sharing permissions and usage restrictions. Metadata is licensed under CC0, allowing unrestricted collection through external queries. All community content is openly licensed.

Ownership
When uploading to community no ownership rights are transferred to the community as well as no licence transfer.

Versioning
Data files can be versioned. Records are not versioned. The uploaded data is archived as a Submission Information Package. Records can be retracted from public view; however, the data files and record are preserved within the system.