cubedata.py generates our lower instance
embed.py contains the terminal embedding
ExTree3.py generates the threshold tree
ExKmeans.py contains the function ExKmeans(X, k, C, numr)
-
X: data set
-
k: number of clusters
-
C: k centers (If not provide, it will run a kmeans++ to compute a center set)
-
numr: number of times to repeat the explainable tree (default = 10)
cube_exkmeans: compare ExKMC algorithm with our algorithm on the cube instance
- To run the ExKMC, you need to install the 'ExKMC'. I could not install ExKMC under Python 3.9, but I can install it under Python 3.8. (I think any version before 3.8 is Ok.)