assess the clusterability of categorical data
To run the testing-based approach, execute demo.m (for Matlab) or demo.R (for R). These scripts call the functions defined in TestCat.m or TestCat.R, and count_attribute_pairs.m or Count_attribute_pairs.R respectively.
For the compared methods, the following scripts run the respective methods:
DistanceMethods.R(2 methods): Dip-dist, Silv-distDimensionalityReductionMethods.R(6 methods): MDS Dip, MDS Silv, tSNE Dip, tSNE Silv, UMAP Dip, UMAP SilvEmbeddingMethods.R(4 methods): PCA Dip, PCA Silv, SPCA Dip, SPCA Silv