The Continue Option allows clustering to be resumed where it left off by starting with the cluster centroid, variance and probability values of the last complete iteration saved in the metadata tables or output tables as requested on the Output Panel. Specifically, if the Continue Option is selected and output tables are specified and exist, the information in the output tables is used to restart processing. If output tables do not exist, then the model in metadata is used to restart processing.
In the case of the Fast K-Means algorithm, however, the Continue Option depends on locating the Cluster Definition table named on the analysis parameters tab, which is effectively the model for this algorithm variation. The Cluster Definition table also is updated for each iteration of the algorithm and can, therefore, provide a degree of recovery.
There is a special case of the Continue Option where using the Fast K-Means algorithm starts processing and, if processing terminates successfully, allows continuing with a Gaussian Mixture Model clustering.
You can request output tables on a Fast K-Means analysis and request the same tables as output tables on a Gaussian Mixture Model analysis with the Continue Option also selected.