ABSTRACT
In this chapter we formulate explicitly the assumptions underlying the complexity analysis introduced in the previous chapter. We discuss their implications and show that their effect is a significant simplification of determining desired performance measures of an algorithm. Many of the assumptions relate to some form of uniformity, be it uniformity in the way operations are counted, uniformity in accessing memory, or uniformity in the validity of mathematical identities. We also reexamine the asymptotic nature of the functions that result from determining complexities. While most of these aspects appear fairly innocuous, their discussion sets up the exploration in Part 2 of whether these assumptions remain valid when designing software based on the analyzed algorithms.
