Floating-Point L-Approximations
نویسندگان
چکیده
Computing good polynomial approximations to usual functions is an important topic for the computer evaluation of those functions. These approximations can be good under several criteria, the most desirable being probably that the relative error is as small as possible in the L sense, i.e. everywhere on the interval under study. In the present paper, we investigate a simpler criterion, the L case. Though finding a best polynomial L-approximation with real coefficients is quite easy, we show that if the coefficients are restricted to be floating point numbers to some precision, the problem becomes a general instance of the CVP problem, and hence is NP-hard. We investigate the practical behaviour of exact and approximate algorithms for this problem. The conclusion is that it is possible in a short amount of time to obtain a relative or absolute best L-approximation. The main applications are for large dimension, as a preliminary step of finding L-approximations and for functions with large variations, for which relative best approximation is by far more interesting than absolute. Key-words: Floating-point arithmetic, efficient approximation, L norm, L norm, lattice basis reduction, closest vector problem. ∗ Université de Saint-Étienne in ria -0 01 19 25 4, v er si on 2 11 D ec 2 00 6 Approximation flottante au sens L Résumé : Calculer de bons polynômes d’approximation pour les fonctions usuelles est un sujet important, par exemple pour obtenir des méthodes performantes d’évaluation de ces fonctions. La qualité desdites approximations peut être mesurée à l’aune de plusieurs critères. Le critère le plus intéressant pour les applications est probablement la minimisation de l’erreur relative au sens L, c’est-à-dire uniformément sur l’intervalle considéré. Dans ce travail, nous étudions un critère plus simple, à savoir la minimisation au sens L. Bien que la recherche de l’optimum à coefficients réels soit une tâche aisée, nous montrons que, si les coefficients sont des nombres flottants de précision fixée, le problème devient une instance générale du problème CVP, et est donc NP-difficile. Nous étudions néanmoins le comportement pratique des algorithmes exacts et approchés pour ce problème. Notre conclusion est qu’il est possible, de façon efficace, d’obtenir la meilleure approximation au sens L, absolu ou relatif. Les principales applications sont la recherche d’approximants en grande dimension, l’utilisation comme étape préliminaire pour trouver des apporximations L, ou pour les fonctions avec de grandes variations, pour lesquelles l’approximation relative est de loin préférable. Mots-clés : Arithmétique flottante, approximation efficace, norme L, norme L, réduction de réseaux, recherche de vecteur le plus proche. in ria -0 01 19 25 4, v er si on 2 11 D ec 2 00 6 Floating-Point L-Approximations 3
منابع مشابه
Exploring Approximations for Floating-Point Arithmetic using UppSAT
We consider the problem of solving floating-point constraints obtained from software verification. We present UppSAT — an new implementation of a systematic approximation refinement framework [24] as an abstract SMT solver. Provided with an approximation and a decision procedure (implemented in an off-the-shelf SMT solver), UppSAT yields an approximating SMT solver. Additionally, UppSAT yieldsi...
متن کاملApproximating floating-point operations to verify numerical programs∗
The verification of programs performing floating-point computations is a key issue in the development of critical software. Reasoning on floatingpoint computations is a tricky task that requires dedicated tools: floatingpoint arithmetic is not correctly handled by solvers over the reals. Several methods have been developed in order to verify programs working with floating-point numbers. Some of...
متن کاملOptimization of Embedded Linux systems without FPU
1 Linux is a popular operating system for embedded systems. Some embedded systems use processors without floating point accelerators (FPA) or floating point units (FPU), in order to satisfy cost and power consumption requirements. This paper describes and compares optimization options for software implementation of floating point operations on systems without FPA or FPU. Our results show that s...
متن کاملRefining Abstract Interpretation-based Approximations with Constraint Solvers
Programs with floating-point computations are tricky to develop because floating-point arithmetic differs from real arithmetic and has many counterintuitive properties. A classical approach to verify such programs consists in estimating the precision of floating-point computations with respect to the same sequence of operations in an idealized semantics of real numbers. Tools like Fluctuat—base...
متن کاملFloating-Point LLL: Theoretical and Practical Aspects
The text-book LLL algorithm can be sped up considerably by replacing the underlying rational arithmetic used for the Gram-Schmidt orthogonalisation by floating-point approximations. We review how this modification has been and is currently implemented, both in theory and in practice. Using floating-point approximations seems to be natural for LLL even from the theoretical point of view: it is t...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2006