K-means algorithm with optional equal cardinality constraint (Pierre David Belanger: https://github.com/pierredavidbelanger/ekmeans) modified and ported to MATLAB.
Input is a matrix where columns are dimensions and rows are data-points.
Outputs are cluster assignments for each data-point and the centroid co-ordinate of each cluster.
Random centroids are seeded in the data and data points are assigned to these centroids on the basis of their euclidian distance. Assignments and centroid positions are iteratively updated. If equal cardinality is specified, whenever a cluster becomes too large points in that cluster that are closest to any other centroid are removed and reassigned to their next closest centroid and all clusters/centroids updated.
If equal cardinality is not achieved in the number of iterations specified, the algorithm is re-initialised from different starting conditions and run again until equal cardinality is achieved.
GUI to view, evaluate and optimise cluster assignments. Can plot in up to 3 dimensions. GUI stores up to 10 cluster splits in memory and user can cycle through/replace clusters with GUI buttons.
Inputs:
- .csv/.txt: matrix where columns are dimensions and rows are data-points.
- .mat: where first variable is matrix where columns are dimensions and rows are data-points.
- .tiff: all pixels are 0 except data-point pixels which are > 0
- matrix: all values are 0 except data-points which are > 0
Input can be provided as workspace variables or selected in file-selection GUI.
Input can be provided as a workspace variable or not provided (select in file-section GUI).
An example of each of these data-types in 2D/3D is provided in this repository.
E.g. to cluster 120 2D data points into equal clusters of 12 from .csv file:
%% NB USERPATH below is the path to your ekmeans repository
yx = load('USERPATH/ekmeans/Test_120points_2D.csv')
num_points_in_cluster = 12;
equal = 1;
[assignments,centroids] = ekmeansGUI(num_points_in_cluster,equal,'input',yx);Scatterplots 2D/3D clusters, connected by lines to their respective centroids and applies an arbitrary colourcode to aid visualisation.