Theorie"en voor het berekenen van algoritmen (met een samenvatting in het Nederlands) PROEFSCHRIFT ter verkrijging van de graad van doctor aan de Rijksuniversiteit te Utrecht op gezag van de Rector Magnificus, Prof. Dr. J.A. van Ginkel ingevolge het besluit van het College van Dekanen

To my parents and uncle Frank

How to decrease labor and improve reliability in the development of efficient implementations of nonnumerical algorithms and labor intensive software is an increasingly important problem as the demand for computer technology shifts from easier applications to more complex algorithmic ones; e.g., optimizing compilers for supercomputers, intricate data structures to implement efficient solutions to operations research problems, search and analysis algorithms in genetic engineering, complex software tools for workstations, design automation, etc. It is also a difficult problem that is not solved by current CASE tools and software management disciplines, which are oriented towards data processing and other applications, where the implementation and a prediction of its resource utilization follow more directly from the specification. Recently, Cai and Paige reported experiments suggesting a way to implement nonnumerical algorithms in C at a programming rate (i.e., source lines per second) t...

Dynamic programming is an important algorithm design technique. It is used for solving problems whose solutions involve recursively solving subproblems that share subsubproblems. While a straightforward recursive program solves common subsubproblems repeatedly and often takes exponential time, a dynamic programming algorithm solves every subsubproblem just once, saves the result, reuses it when the subsubproblem is encountered again, and takes polynomial time. This paper describes a systematic method for transforming programs written as straightforward recursions into programs that use dynamic programming. The method extends the original program to cache all possibly computed values, incrementalizes the extended program with respect to an input increment to use and maintain all cached results, prunes out cached results that are not used in the incremental computation, and uses the resulting incremental program to form an optimized new program. Incrementalization statically exploits semantics of both control structures and data structures and maintains as invariants equalities characterizing cached results. The principle underlying incrementalization is general for achieving drastic program speedups. Compared with previous methods that perform memoization or tabulation, the method based on incrementalization is more powerful and systematic. It has been implemented and applied to numerous problems and succeeded on all of them. 1

This paper describes a method for transforming any given set of Datalog rules into an efficient specialized implementation with guaranteed worst-case time and space complexities, and for computing the complexities from the rules. The running time is optimal in the sense that only useful combinations of facts that lead to all hypotheses of a rule being simultaneously true are considered, and each such combination is considered exactly once. The associated space usage is optimal in that it is the minimum space needed for such consideration modulo scheduling optimizations that may eliminate some summands in the space usage formula. The transformation is based on a general method for algorithm design that exploits fixed-point computation, incremental maintenance of invariants, and combinations of indexed and linked data structures. We apply the method to a number of analysis problems, some with improved algorithm complexities and all with greatly improved algorithm understanding and greatly simplified complexity analysis.

Object abstraction supports the separation of what operations are provided by systems and components from how the operations are implemented, and is essential in enabling the construction of complex systems from components. Unfortunately, clear and modular implementations have poor performance when expensive query operations are repeated, while efficient implementations that incrementally maintain these query results are much more difficult to develop and to understand, because the code blows up significantly, and is no longer clear or modular. This paper describes a powerful and systematic method that first allows the “what ” of each component to be specified in a clear and modular fashion and implemented straightforwardly in an object-oriented language; then analyzes the queries and updates, across object abstraction, in the straightforward implementation; and finally derives the sophisticated and efficient “how ” of each component by incrementally maintaining the results of repeated expensive queries with respect to updates to their parameters. Our implementation and experimental results for example applications in query optimization, role-based access control, etc. demonstrate the effectiveness and benefit of the method.

This paper briefly surveys what transformational programming is about, and how to make progress in the field. A program transformation is a meaning-preserving mapping defined on a programming language. Transformational programming is a program development methodology in which an implementation I is obtained from a specification S by applying a sequence T k :::T 1 of transformations to S. If S and each transformation T i applied to successive implementations T i\Gamma1 :::T 1 S of S are proved correct for

Given are a directed edge-labelled graph G with a distinguished node n0, and a regular expression P which may contain variables. We wish to compute all substitutions φ (of symbols for variables), together with all nodes n such that all paths n0 → n are in φ(P). We derive an algorithm for this problem using relational algebra, and show how it may be implemented in Prolog. The motivation for the problem derives from a declarative framework for specifying compiler optimisations. 1 Bob Paige and IFIP WG 2.1 Bob Paige was a long-standing member of IFIP Working Group 2.1 on Algorithmic Languages and Calculi. In recent years, the main aim of this group has been to investigate the derivation of algorithms from specifications by program transformation. Already in the mid-eighties, Bob was way ahead of the pack: instead of applying transformational techniques to well-worn examples, he was applying his theories of program transformation to new problems, and discovering new algorithms [16, 48, 52]. The secret of his success lay partly in his insistence on the study of general algorithm design strategies (in particular

This paper presents program analyses and transformations for strengthening invariants for the purpose of efficient computation. Finding the stronger invariants corresponds to discovering a general class of auxiliary information for any incremental computation problem. Combining the techniques with previous techniques for caching intermediate results, we obtain a systematic approach that transforms non-incremental programs into ecient incremental programs that use and maintain useful auxiliary information as well as useful intermediate results. The use of auxiliary information allows us to achieve a greater degree of incrementality than otherwise possible. Applications of the approach include strength reduction in optimizing compilers and finite differencing in transformational programming.

This paper describes the precise specification, design, analysis, implementation, and measurements of an efficient algorithm for solving regular tree grammar based constraints. The particular constraints are for dead-code elimination on recursive data, but the method used for the algorithm design and complexity analysis is general and applies to other program analysis problems as well. The method is centered around Paige's finite differencing, i.e., computing expensive set expressions incrementally, and allows the algorithm to be derived and analyzed formally and implemented easily. We study higherlevel transformations that make the derived algorithm concise and allow its complexity to be analyzed accurately. Although a rough analysis shows that the worst-case time complexity is cubic in program size, an accurate analysis shows that it is linear in the number of live program points and in other parameters, including mainly the arity of data constructors and the number of selector applications into whose arguments the value constructed at a program point might flow. These parameters explain the performance of the analysis in practice. Our implementation also runs two to ten times as fast as a previous implementation of an informally designed algorithm.