A "Lazy" Solution to Imprecision in Computational Geometry (1993)
| Venue: | In Proceedings of the 5th Canadian Conference on Computational Geometry |
| Citations: | 14 - 2 self |
BibTeX
@INPROCEEDINGS{Benouamer93a"lazy",
author = {M. O. Benouamer and P. Jaillon and D. Michelucci and J-m. Moreau},
title = {A "Lazy" Solution to Imprecision in Computational Geometry},
booktitle = {In Proceedings of the 5th Canadian Conference on Computational Geometry},
year = {1993},
pages = {73--78}
}
Bookmark
 |
 |
 |
 |
 |
 |
OpenURL
Abstract
Making safe and consistent decisions is essential to geometric algorithms. Quite a few solutions to this problem have been suggested in the recent years, but they generally ask that drastic changes be made to algorithms. A simple and efficient paradigm is suggested here, which enables programmers to forget about precision issues altogether, whatever algorithm they intend to implement. The paper describes a lazy exact arithmetic library (LEA) based on this paradigm and its operation in a typical situation from Computational Geometry. 1 Introduction Lazy languages (ML, Miranda) emerged a few years ago, introducing the strong paradigm of laziness: "Why should a quantity be evaluated if it is never going to be used later?" The same applies to precision: "Why should a quantity be computed exactly if it is never involved in conflicting issues later on?" Of course, if brutally implemented, such an observation would be fatal in situations where some exact information -- that was available bu...
