Design and applications of variational representations

Variability is an important and widely studied topic across domains such as version control, software product lines, and metaprogramming. This thesis presents an investigation into the process of systematically adding variability to data structures and programs, leading to guidelines for variational data structures and implications for programs that create, manipulate, and otherwise consume them. In particular, it focuses on the tradeoffs between expressiveness and efficiency that must be managed when designing these structures and programs. While more general and expressive variational programs can simplify the work programmers need to do, they are inevitably less efficient than approaches tailored to specific scenarios and domains.

This means that these approaches to variability are more suitable for a certain class of applications than others. While efficiently consuming variational data remains a challenge, concentrating on domains that focus on generating variational artifacts has been effective. I demonstrate this directly with two example application domains, namely variational pictures and variational data visualization.

Major Advisor: Martin Erwig
Committee: Eric Walkingshaw
Committee: Eugene Zhang
Committee: Margaret Burnett
GCR: Yelda Turkan