Once the components of one complex have been clustered, and the components of a
second complex have been clustered, it remains to find the correspondence
between the classes. This is another clustering step, using one representative
from each cluster of each complex. From the resulting clusters, singletons are
removed. The result is a mapping from the classes in one complex to those in
another complex.
This is a prerequisite for any component mapping between complexes.
In the case where we have a lookup to classify a component, we can skip both
clustering steps.
Original issue reported on code.google.com by chad.a.davis@gmail.com on 27 May 2011 at 7:33
Original issue reported on code.google.com by
chad.a.davis@gmail.com
on 27 May 2011 at 7:33