Computation tends to produce discrete, self-similar parts and aggregate them together, a common concern in recent projects that we think of as computational. How can we think about using the computational techniques available to us, but in a different way than they have been studied? This thesis is an exploration about the usefulness of these techniques at a greater diversity of scale than what is seen in digital fabrication or product sets. Reversing the logic of aggregation in order to produce wholes that subdivide into parts, it poses the experiment of producing a Closed Whole with Non-Standard Parts produced as an Object Set. It takes as a platform the issue of a housing model in the city of Houston. The work begins as a formal investigation to explore what the potentials of these techniques might be, and posits that we can respond to an architectural condition at the scale of part-to-whole relationships and to an urban condition at the scale of the object set, both intrinsic to a singular methodology.