ParaKMeans: Implementation of a parallelized K-means algorithm suitable for general laboratory use-3

He number of compute nodes used in the analysis. The bar graphs at the bottom of each plot illustrate the number of compute nodes where one finds statistically significant increases in speed. The p values presented are for tests of differences between the number of compute nodes for a given number of genes or clusters. The effect of the interaction between the number of genes and number of compute nodes on the speed of execution. The effect of the interaction between the number of clusters and number of compute nodes on the speed of execution.<p><b>Copyright information:</b></p><p>Taken from "ParaKMeans: Implementation of a parallelized K-means algorithm suitable for general laboratory use"</p><p>http://www.biomedcentral.com/1471-2105/9/200</p><p>BMC Bioinformatics 2008;9():200-200.</p><p>Published online 16 Apr 2008</p><p>PMCID:PMC2375128.</p><p>