A Computer Oriented Approach to Get Sharp Reliable Error Bounds

We present interval methods to get reliable a priori error bounds for the machine evaluation of algorithms implementing some mathematical expression. The term expression not only means simple arithmetical expressions but also more complex program parts including loops or recursive structures (e.g. a complete elementary function routine).We sketch a method that can be used to get an upper bound for the approximation error of a polynomial or a rational approximation. We also discuss a method to compute worst case a priori error estimates for arbitrary IEEE double floating-point computations. Our theoretical results lead to reliable and easy to use public domain software tools. The application of these tools to an accurate table method shows that error bounds of high quality can be derived.     

Least significant bit evaluation of arithmetic expressions in single-precision

Single-precision floatingpoint computations may yield an arbitrary false result due to cancellation and rounding errors. This is true even for very simple, structured arithmetic expressions such as Horner's scheme for polynomial evaluation. A simple procedure will be presented for fast calculation of the value of an arithmetic expression to least significant bit accuracy in single precision computation. For this purpose in addition to the floating-point arithmetic only a precise scalar product (cf. [2]) is required. If the initial floatingpoint approximation is not too bad, the computing time of the new algorithm is approximately the same as for usual floating-point computation. If not, the essential progress of the presented algorithm is that the inaccurate approximation is recognized and corrected. The algorithm achieves high accuracy, i.e. between the left and the right bound of the result there is at most one more floating-point number. A rigorous estimation of all rounding errors introduced by floating-point arithmetic is given for general triangular linear systems. The theorem is applied to the evaluation of arithmetic expressions.     

ESOLID—a system for exact boundary evaluation

We present a system, ESOLID, that performs exact boundary evaluation of low-degree curved solids in reasonable amounts of time. ESOLID performs accurate Boolean operations using exact representations and exact computations throughout. The demands of exact computation require a different set of algorithms and efficiency improvements than those found in a traditional inexact floating-point based modeler. We describe the system architecture, representations, and issues in implementing the algorithms. We also describe a number of techniques that increase the efficiency of the system based on lazy evaluation, use of floating-point filters, arbitrary floating-point arithmetic with error bounds, and lower-dimensional formulation of subproblems.ESOLID has been used for boundary evaluation of many complex solids. These include both synthetic datasets and parts of a Bradley Fighting Vehicle designed using the BRL-CAD solid modeling system. It is shown that ESOLID can correctly evaluate the boundary of solids that are very hard to compute using a fixed-precision floating-point modeler. In terms of performance, it is about an order of magnitude slower as compared to a floating-point boundary evaluation system on most cases.     

Formal Proofs of Rounding Error Bounds

Floating-point arithmetic is a very efficient solution to perform computations in the real field. However, it induces rounding errors making results computed in floating-point differ from what would be computed with reals. Although numerical analysis gives tools to bound such differences, the proofs involved can be painful, hence error prone. We thus investigate the ability of a proof assistant like Coq to mechanically check such proofs. We demonstrate two different results involving matrices, which are pervasive among numerical algorithms, and show that a large part of the development effort can be shared between them.     

Strict optimal a posteriori error and residual bounds for Gaussian elimination in floating-point arithmetic

Exact representations of errors and residuals of approximate solutions of linear algebraic systems under data perturbations and rounding errors of a floating-point arithmetic are established from which strict optimal a posteriori error and residual bounds are obtained. These bounds are formulated by means of a posteriori error and residual condition numbers. Condition numbers, error and residual bounds can be computed completely in the range of nonnegative numbers using the arithmetic operations+, x, / only. It is shown that computations in this range are numerically very stable. The general results are applied to a series of numerical examples.     

A framework for dynamic implicit curve approximation by an irregular discrete approach

The approximation of implicit planar curves by line segments is a very classical problem. Many algorithms use interval analysis to approximate this curve, and to handle the topology of the final reconstruction. In this article, we use discrete geometry tools to build an original geometrical and topological representation of the implicit curve. The polygonal approximation contains few segments, and the Reeb graph permits to sum up efficiently the shape and the topology of the curve. Furthermore, we propose two algorithms to process local cells refinement and local cells grouping schemes. We illustrate these schemes with a global system that efficiently handles manual or automatic fast updates on the global reconstruction, by considering topological or geometrical constraints. We also compare the speed and the quality of our approach with two classical methods.     

Review on stochastic approach to round-off error analysis and its applications

Because of round-off error propagation in floating-point arithmetic, any result provided by a computer always contains an error. The Permutation-Perturbation method, also known under the CESTAC (Contrôle et Estimation Stochastique des Arrondis de Calcul) method is a precious tool: (i) for evaluating the accuracy of results provided by direct algorithms performed by a computer, (ii) for breaking off correctly any iterative process, and for evaluating the accuracy of the results, and (iii) for determining the optimal step or the optimal mesh in the approximate algorithms. A review of this method and of these applications is presented in this paper.     

Low-degree approximation of high-degree B-spline surfaces

In this paper, a method for approximate conversion of high degree Bezier and B-spline surfaces to lower degree representations is presented to facilitate the exchange of surface geometry between different geometric modeling systems. Building on previous work on curve approximation, the method uses adaptive sampling to compute approximation error and lofting of isoparametric curves to produce the approximating surface. In addition, a bound for the approximation accuracy is computed using convex hulls.     

Practical methods for evaluating the accuracy of the eigenelements of a symmetric matrix

The results of the Householder and QL algorithms for determining the eigenelements of a symmetric matrix, provided by a computer, always contain the errors resulting from floating-point arithmetic round-off error propagation. The Permutation-Perturbation method is a very efficient practical method for evaluating these errors and consequently for estimating the exact significant figures of the eigenelements. But, in the cases of: eigenvalues very close to zero, eigenvalues of widely varying range, and multiple eigenvalues, the Permutation-Perturbation method is not complete. In this paper we propose an algorithm which is able to complete this method.     

基于遗传算法的B样条曲线和Bézier曲线的最小二乘拟合

考虑用B样条曲线拟合平面有序数据使得最小二乘拟合误差最小.一般有两种考虑,一种是保持B样条基函数的节点不变,选择参数使得拟合较优.参数的选择方法包括均匀取值、累加弦长法、centripetal model、Gauss-Newton迭代法等.另一种则是先确定好参数值(一般用累加弦长法),然后再用.某一算法计算出节点,使得拟合较优.同时把两者统一考虑,用遗传算法同时求出参数、节点使得拟合在最小二乘误差意义下最优.与Gauss-Newton迭代法、Piegl算法相比,本方法具有较好的鲁棒性(拟合曲线与初始值无关)、较高的精度及控制顶点少等优点.实验结果说明采用遗传算法得到的曲线逼近效果更好.用遗传算法对Bezier曲线拟合平面有序数据也进行了研究.     

Trade-offs of certified fixed-point code synthesis for linear algebra basic blocks

In embedded systems, efficient implementations of numerical algorithms typically use the fixed-point arithmetic rather than the standardized and costly floating-point arithmetic. But, fixed-point developers face two difficulties: First, writing fixed-point codes is tedious and error prone. Second, the low dynamic range of fixed-point numbers leads to the persistent belief that fixed-point computations are inherently inaccurate. In this article, we address these two limitations by introducing a methodology to design and implement tools that synthesize fixed-point programs. To strengthen the user’s confidence in the synthesized code, analytic methods are presented to automatically assert its numerical quality. Furthermore, we use this framework to generate fixed-point code for linear algebra basic blocks such as matrix multiplication and inversion. For example, the former task involves trade-offs such as choosing to maximize the code’s accuracy or minimize its size. For the two cases of matrix multiplication and inversion, we describe, implement, and experiment with several algorithms to find trade-offs between the conflicting goals.     

Bézier样条曲线的近似弧长参数化方法

提出了Bézier样条曲线近似弧长参数化的方法及相应的算法。通过求出曲线近似二分之一弧长的点及其相应的参数值,可将曲线分割为两条Bézier样条曲线。这两条曲线的弧长近似相等,因此让它们带有相同的权1。对新生成的Bézier样条曲线不断重复上述工作,最终得到一条由多条Bézier样条曲线所构成的新的曲线。将这多条Bézier样条曲线合并为一条Bézier样条曲线,进而通过节点插入技术将其转化为B样条形式的曲线以便得到全局参数,其中各段Bézier曲线在全局参数域中所占子区间的长度与它们所具有的权成比例,这样便得到一条近似弧长参数化曲线。     

A Verified ODE Solver and the Lorenz Attractor

A rigorous numerical algorithm, formally verified with Isabelle/HOL, is used to certify the computations that Tucker used to prove chaos for the Lorenz attractor. The verification is based on a formalization of a diverse variety of mathematics and algorithms. Formalized mathematics include ordinary differential equations and Poincaré maps. Algorithms include low level approximation schemes based on Runge–Kutta methods and affine arithmetic. On a high level, reachability analysis is guided by static hybridization and adaptive step-size control and splitting. The algorithms are systematically refined towards an implementation that can be executed on Tucker’s original input data.     

Numerical reproducibility for the parallel reduction on multi- and many-core architectures

On modern multi-core, many-core, and heterogeneous architectures, floating-point computations, especially reductions, may become non-deterministic and, therefore, non-reproducible mainly due to the non-associativity of floating-point operations. We introduce an approach to compute the correctly rounded sums of large floating-point vectors accurately and efficiently, achieving deterministic results by construction. Our multi-level algorithm consists of two main stages: first, a filtering stage that relies on fast vectorized floating-point expansion; second, an accumulation stage based on superaccumulators in a high-radix carry-save representation. We present implementations on recent Intel desktop and server processors, Intel Xeon Phi co-processors, and both AMD and NVIDIA GPUs. We show that numerical reproducibility and bit-perfect accuracy can be achieved at no additional cost for large sums that have dynamic ranges of up to 90 orders of magnitude by leveraging arithmetic units that are left underused by standard reduction algorithms.     

Motivations for an Arbitrary Precision Interval Arithmetic and the MPFI Library

This paper justifies why an arbitrary precision interval arithmetic is needed. To provide accurate results, interval computations require small input intervals; this explains why bisection is so often employed in interval algorithms. The MPFI library has been built in order to fulfill this need. Indeed, no existing library met the required specifications. The main features of this library are briefly given and a comparison with a fixed-precision interval arithmetic, on a specific problem, is presented. It shows that the overhead due to the multiple precision is completely acceptable. Eventually, some applications based on MPFI are given: robotics, isolation of polynomial real roots (by an algorithm combining symbolic and numerical computations) and approximation of real roots with arbitrary accuracy.This work was done while N. Revol was a member of the ANO Laboratory, University of Lille, France, on sabbatical leave within the Arenaire project.This work was done while F. Rouillier belonged to the Spaces project, LORIA and LIP6, France.     

Managing uncertainty and discontinuous condition numbers in geometric computation

We consider operations on subdivision surfaces under the strict robustness requirement that these floating-point computations return an object with the same topological form as the true solution. The problems involved may however be ill-conditioned, and defined in terms of uncertain data, and even supplementary interval arithmetic may not ensure robustness. Trapping mechanisms are therefore proposed to resolve this difficulty.     

近似弧长参数化Bézier曲线的最佳逼近

考虑近似弧长参数化Bézier曲线的逼近问题.当获得Bézier曲线的一个近似弧长参数化之后,这种参数化只能达到C0-连续性.为了增加其参数连续性,利用其带有端点约束的关于L2-模的最佳逼近以得到具有C2-连续性的Bézier样条曲线.实验证明,这种逼近的效果是十分理想的.     

曲线的整数型生成算法

本文提出了一个用光栅显示器或数字化绘图仪等显示设备中选择曲线上最佳点的通过算法，该算法由几部分组成，分别对应曲线的不同走向段，其最大的特点是可以根据实际曲线的走向，在算法的各部分实现自动跳动，由此算法可生成所有常用曲线，本文给出Ｂｅｚｉｅｒ曲线和Ｂ样条曲线的生成算法，这些算法选择距离实际曲线最近的网格点，并且只有整数运算。     

区间Ball曲线的边界及降阶

讨论了n次区间Ball曲线的边界的构成;同时通过讨论区间多项式的降阶,利用线性规划法及最佳一致逼近法,给出了区间Ball曲线的的降阶算法．若利用线性规划法得到的区间曲线不能达到预期的误差,则可以结合细分的技术实现．     

增量极坐标编码的贝赛尔曲线智能优化算法

针对基于统计的隶属度函数确定方法进行了改进,使用贝塞尔曲线作为隶属度函数的上升或下降沿,使隶属度函数可以经过统计结果规定的任意中间点。使用新的增量极坐标编码对贝塞尔曲线控制点进行表达,解决了传统贝塞尔曲线优化中的控制点约束问题。采用差分进化算法对贝塞尔曲线控制点进行优化,可智能拟合经过任意点的最佳贝塞尔曲线。算法可扩展到任意阶贝塞尔曲线,所得隶属度函数较非贝塞尔曲线方法更为合理。