{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

johnalexpres

# johnalexpres - everything you wanted to know about...

This preview shows pages 1–10. Sign up to view the full content.

everything you wanted to know about computers* John Alex *but were too afraid to ask

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Overview • Functions as representation • Function optimization methods, issues – Steepest Descent – Simulated Annealing – Genetic Algorithms • Analysis of shape grammars • Possibilities
Ode to Functions • Math is based on them • Computers are based on them • Very general representation: a mapping • Helpful as intermediate object too – aid to formalization, rigor • Limited – only maps numbers to numbers – is mapping it ?

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Functions y = f(x) • x,y: vector of parameters (‘parametric’?) • “Form function” – Vertices = f(dimensions, key pts, etc) • “Fitness function” – Quality = f(vertices)
Functions y = f(x) • x,y are each a vector of parameters • Each parameter can be either – discrete (combinatorial): 0, 1, 2, 3, 4 – continuous: 0-4

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Functions y = f(x) • such that c(x) = 0 • Constraints: valid parameter combinations
Trouble with Functions in Design • Pre-Optimization questions: – how to define a useful form function? • Vertices = f(dimensions, key pts, etc) – how to define a useful fitness function? • fitness = f(geometry only)? – generality vs. specificity – myth: computer functions can be random – myth: designers’ functions are random

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Trouble with Functions in Design • Optimization question: – how to find the most fit form? • Pre, mid, post-optimization question: – how to handle emergence? • changing form, fitness functions during design • changing question in middle of trying to answer it
Architectural Function Optimization • Vertices = f(model parameters) • Quality = f(vertices) • Quality = f(model parameters) Optimization = vary model parameters to maximize goodness

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}