Each folder represents a branch of the repository containing experiments on a particular data set.
Experiments were performed in Jupyter Notebooks.
The experiments from the main paper are located in the < understanding-data-values-main> folder. 

Each repository is structured in the following way: 

	The folder <utils> contains all our code. This includes the synthetic data set, plotting methods and methods for data valuation. 
		For DVRL and Shapley we borrow code from the original papers. 

	The folder <e0_exploration> contains some baseline experiments we performed.

	The folder <e1_ood > contains experiments with out-of-distribution methods.

	The folder < e2_synthetic_data> contains the main experiments.

Additionally, the <db_change.gif> is a visual explanation of our experiments.
