Spectral Clustering

Implementation of Spectral Clustering in Apache Spark.

Dataset used is synthetic data, generated on-the-fly using random number generators (specifically, the scikit-learn samples generators); they don’t represent any “real” data

Used Matplot Library for plotting the clusters

How to Run

$ spark-submit ~/absolute/path/to/the/directory/spectral_clustering.py 3 10 1.0 a5_data/blobs.txt

• Argument 1 is the number of clusters
• Argument 2 is the upper bound
• Argument 3 is the value of gamma

Note:- spark-submit should be in path