Using Rhino3D and parametric plugins, a procedural city model was developed based on three primary inputs: street network, building geometry, and density allocation. The street network defines block boundaries and overall urban structure, while buildings are generated parametrically to fit within each block. Density is controlled through attraction points that distribute program intensity across the site, allowing building height, footprint, and massing to adapt according to location.

With input 1 and 2, the building geometries are allocated to fit the street grid shape with different density allocation points. The density allocation points determines which area is downtown and which is suburb, providing a more contextual elements to the cities generated. With 3 iterations of each input, 3 x 3 x 3 creates a total 27 numbers of cities generated.

Grasshopper Script for City Generation:

The solar analysis shows clear pattern where south facing facade receives the most light, while there are some surfaces with high obstruction receiving less solar radiation. This shows that for building's solar exposure, usually orientation is more important than surrounding.

In case of view analysis, the general pattern is that higher floor is always better. Compared to solar analysis, the view depends much more on the surrounding, where the existence of higher buildings affects view much more than the height or the orientation of the building.