Clustering_toolbox.zip_clustering_zip

1. use clustering_pipeline_masterscript.m to run the analysis. all other functions are called from this script. 2. to install the toolbox, cd to the 'ERP_clustering_toolbox' directory and run: addpath(genpath(pwd)); (or evaluate line 11 of clustering_pipeline_masterscript.m) 3. enter path to input dataset as a string, to 'infname' in line 14 of clustering_pipeline_masterscript.m 4. for emd_cutoff.m function, maxf parameter should be set high (~50), since the method to find the knee of the function needs enough points to fit two straight lines. NOTE: the algorithm in emd_cutoff.m has not yet been validated on simulated data. this is why it has been commented out in clustering_pipeline_manuscript.m. DO NOT use it until further notice. instead, set co~10. 5. results are saved after EMD-denoising (emd_den_dataset.mat), estimation of clustering stability (stability.mat), estimation of clustering stability (stability_nulldist.mat), calculating of initial cluster centroids using GA (init_centroids.mat). any of these datasets can be loaded and analysis started from that point, if so required. if using this code for your analysis, please cite: Williams NJ, Nasuto SJ, Saddy, JD. Method for exploratory cluster analysis and visualisation of single-trial ERP ensembles. Journal of Neuroscience Methods 250, 22-33.