SPSS- cross validation of cluster analysis

In SPSS, I conducted Ward’s method of hierarchical cluster analysis based on squared Euclidean distances followed by k-means nonhierarchical clustering. I am now in the process of validating my final cluster solution by using a cross-validation of a random 50% of the sample.

I was able to create new variables that randomly split the sample, but I am unsure where to go from here. Is it possible to select only half the data sample to carry out cluster analysis? If so, how?

Let me know if you have any clarification questions--I'm just generally confused by cross-validation.