Dynamic Symbolic Optimisation for Relation-Algebraic Programming in Haskell

Wolfram Kahl

pp. 92-99 in Dongming Wang, Zhiming Zheng (eds.) Proc. First International Conference on Mathematical Aspects of Computer and Information Systems, MACIS 2006, Beijing, Beihang University, 2006.

(.bib)

Abstract

When using efficient BDD-based algorithms for relation-algebraic calculations, the efficiency of evaluating equivalent expressions with apparently similar structure may differ widely, and in unexpected ways, depending on the precise behaviour of the heuristics used in the underlying BDD-operations.

After motivating the relation-algebraic programming style and highlighting these efficiency problems, we describe the set of strategies we are currently pursuing to arrive at a ``self-optimising'' implementation that frees the user of the relation-algebraic programming interface from having to consider efficiency concerns.


Wolfram Kahl