>>> from mlpack import kmeans
This program performs K-Means clustering on the given dataset. It can return the learned cluster assignments, and the centroids of the clusters. Empty clusters are not allowed by default; when a cluster becomes empty, the point furthest from the centroid of the cluster with maximum variance is taken to fill that cluster.
Optionally, the Bradley and Fayyad approach ("Refining initial points for k-means clustering", 1998) can be used to select initial points by specifying the 'refined_start' parameter. This approach works by taking random samplings of the dataset; to specify the number of samplings, the 'samplings' parameter is used, and to specify the percentage of the dataset to be used in each sample, the 'percentage' parameter is used (it should be a value between 0.0 and 1.0).
There are several options available for the algorithm used for each Lloyd iteration, specified with the 'algorithm' option. The standard O(kN) approach can be used ('naive'). Other options include the Pelleg-Moore tree-based algorithm ('pelleg-moore'), Elkan's triangle-inequality based algorithm ('elkan'), Hamerly's modification to Elkan's algorithm ('hamerly'), the dual-tree k-means algorithm ('dualtree'), and the dual-tree k-means algorithm using the cover tree ('dualtree-covertree').
The behavior for when an empty cluster is encountered can be modified with the 'allow_empty_clusters' option. When this option is specified and there is a cluster owning no points at the end of an iteration, that cluster's centroid will simply remain in its position from the previous iteration. If the 'kill_empty_clusters' option is specified, then when a cluster owns no points at the end of an iteration, the cluster centroid is simply filled with DBL_MAX, killing it and effectively reducing k for the rest of the computation. Note that the default option when neither empty cluster option is specified can be time-consuming to calculate; therefore, specifying either of these parameters will often accelerate runtime.
Initial clustering assignments may be specified using the 'initial_centroids' parameter, and the maximum number of iterations may be specified with the 'max_iterations' parameter.
As an example, to use Hamerly's algorithm to perform k-means clustering with k=10 on the dataset 'data', saving the centroids to 'centroids' and the assignments for each point to 'assignments', the following command could be used:
>>> output = kmeans(input=data, clusters=10)
>>> assignments = output['output']
>>> centroids = output['centroid']
To run k-means on that same dataset with initial centroids specified in 'initial' with a maximum of 500 iterations, storing the output centroids in 'final' the following command may be used:
>>> output = kmeans(input=data, initial_centroids=initial, clusters=10,
>>> final = output['centroid']
- clusters (int): [required] Number of clusters to find (0 autodetects from initial centroids). Default value 0.
- input (numpy matrix or arraylike, float dtype): [required] Input dataset to perform clustering on.
- algorithm (string): Algorithm to use for the Lloyd iteration ('naive', 'pelleg-moore', 'elkan', 'hamerly', 'dualtree', or 'dualtree-covertree'). Default value naive.
- allow_empty_clusters (bool): Allow empty clusters to be persist.
- copy_all_inputs (bool): If specified, all input parameters will be deep copied before the method is run. This is useful for debugging problems where the input parameters are being modified by the algorithm, but can slow down the code.
- in_place (bool): If specified, a column containing the learned cluster assignments will be added to the input dataset file. In this case, --output_file is overridden. (Do not use in Python.)
- initial_centroids (numpy matrix or arraylike, float dtype): Start with the specified initial centroids.
- kill_empty_clusters (bool): Remove empty clusters when they occur.
- labels_only (bool): Only output labels into output file.
- max_iterations (int): Maximum number of iterations before k-means terminates. Default value 1000.
- percentage (float): Percentage of dataset to use for each refined start sampling (use when --refined_start is specified). Default value 0.02.
- refined_start (bool): Use the refined initial point strategy by Bradley and Fayyad to choose initial points.
- samplings (int): Number of samplings to perform for refined start (use when --refined_start is specified). Default value 100.
- seed (int): Random seed. If 0, 'std::time(NULL)' is used. Default value 0.
- verbose (bool): Display informational messages and the full list of parameters and timers at the end of execution.
The return value from the binding is a dict containing the following elements:
- centroid (numpy matrix, float dtype): If specified, the centroids of each cluster will be written to the given file.
- output (numpy matrix, float dtype): Matrix to store output labels or labeled data to.