Implemented a choice for the kernels used in the calculation of local density. There are three default kernels (flat, exponential and gaussian) and custom kernels can be created as well by passing a function object. This kernels can be selected from python using the chooseKernel function
Changed the data members of the clusterer class from lists to np.arrays for better performance
Removed printouts of the partial execution times, used in the initial testing phases
Added more user customization for the input and the cluster plotter
Added choice of the size of the space where the blobs are created (useful when we want to create a lot of blobs)
Added verbose option for runCLUE, which prints the number of clusters found and the execution time
Implemented test for the execution times in the continuous integration
Added dictionaries to the accepted input data types
List of changes for version 1.3.0: