基于二元连分式的散乱数据递推插值格式

本文首先基于新的非张量积型偏逆差商递推算法,分别构造奇数与偶数个插值节点上的二元连分式散乱数据插值格式,进而得到被插函数与二元连分式间的恒等式.接着,利用连分式三项递推关系式,提出特征定理来研究插值连分式的分子分母次数.然后,数值算例表明新的递推格式可行有效,同时,通过比较二元Thiele型插值连分式的分子分母次数,发现新的二元插值连分式的分子分母次数较低,这主要归功于节省了冗余的插值节点. 最后,计算此有理函数插值所需要的四则运算次数少于计算径向基函数插值.     

关于广义逆的向量连分式插值样条

本文首次引入了关于广义逆的向量有理插值样条的概念．这类插值样条具有Thiele型连分式的截断分式的表现形式．在它的构造过程中,不必用到连分式的三项递推关系,本文得到的新的有效的系数算法具有递推运算的特点．存在性的一个充分条件得以建立．包括唯一性在内的有关插值问题的某些结果得到证明．最后,本文给出了一个精确的插值误差公式．     

二元插值的一般框架(英)

本文通过引进多参数建立了二元插值的一般框架.这样,许多著名的经典插值格式,如Newton插值、分叉连分式插值、对称连分式插值等均可视为本文的特殊情形.     

二元插值的一般框架

本文通过引进多参数建立了二元插值的一般框架.这样,许多著名的经典插值格式,如Newton插值、分叉连分式插值、对称连分式插值等均可视为本文的特殊情形．     

二元向量值有理插值存在性的一种判别方法

文章给出了对于矩形网格上基于二元Newton插值公式的二元向量值有理插值存在性的充要条件.在存在的情况下,建立了具有显式表达式的不同于向量连分式的二元向量值有理插值函数,并且这种方法具有承袭性.最后给出的实例说明了这种算法的有效性.     

Thiele重心型矩阵值混合有理插值算法

利用Samelson型矩阵广义逆,构造了一种基于Thiele型连分式插值与重心有理插值的相结合的二元矩阵值混合有理插值格式,这种新的混合矩阵值有理插值函数继承了连分式插值和重心插值的优点,它的表达式简单,计算方便,数值稳定性好.该算法满足有理插值问题所给的插值条件,同时给出了误差估计分析.最后用数值算例验证了插值算法的有效性.     

关于Newton—Thiele型二元有理插值的存在性问题

基于均差的牛顿插值多项式可以递归地实现对待插值函数的多项式逼近,而Thiele型插值连分式可以构造给定节点上的有理函数。将两者结合可以得到Newton-Thiele型二元有理插值（NTRI）算法,本文解决了NTRI算法的存在性问题,并有数值例子加以说明。     

二元矩阵有理插值函数的构造

一般构造矩阵值有理函数的方法是利用连分式给出的,其算法的可行性不易预知,且计算量大.本文对于二元矩阵值有理插值的计算,通过引入多个参数,定义一对二元多项式:代数多项式和矩阵多项式,利用两多项式相等的充分必要条件通过求解线性方程组确定参数,并由此给出了矩阵值有理插值公式.该公式简单,具有广阔的应用前景.     

基于连分式与Newton-Padé逼近的数值积分

首先利用Newton-Pade表中部分序列推导出连分式,提出逆差商算法,算出关于高阶导数与高阶差商的连分式插值余项.接着,构造基于此类连分式的有理求积公式与相应的复化求积公式,算出相应的求积余项,研究表明,在一定条件下,求积公式序列一致收敛于积分真值.然后,为保证连分式计算顺利进行,研究连分式分母非0的充分条件.最后,若干数值算例表明,对某些函数采用新提出的复化有理求积公式计算数值积分,所得结果优于采用Simpson公式.     

切触有理插值函数的新算法

切触有理插值函数的算法大都是基于连分式进行的,其算法的可行性大都是有条件的,且有理函数次数较高,计算量较大.文章利用拉格朗日插值的性质和分段组合的方法,给出了一种新的切触有理插值算法,并给出误差估计且将其推广到向量值切触有理插值情形.较之其他算法,具有有理函数次数较低、计算量较小、算法无条件性、无极点、满足高阶导数插值条件等优点.     

Newton-Thiele's rational interpolants

It is well known that Newton's interpolation polynomial is based on divided differences which produce useful intermediate results and allow one to compute the polynomial recursively. Thiele's interpolating continued fraction is aimed at building a rational function which interpolates the given support points. It is interesting to notice that Newton's interpolation polynomials and Thiele's interpolating continued fractions can be incorporated in tensor‐product‐like manner to yield four kinds of bivariate interpolation schemes. Among them are classical bivariate Newton's interpolation polynomials which are purely linear interpolants, branched continued fractions which are purely nonlinear interpolants and have been studied by Chaffy, Cuyt and Verdonk, Kuchminska, Siemaszko and many other authors, and Thiele-Newton's bivariate interpolating continued fractions which are investigated in another paper by one of the authors. In this paper, emphasis is put on the study of Newton-Thiele's bivariate rational interpolants. By introducing so‐called blending differences which look partially like divided differences and partially like inverse differences, we give a recursive algorithm accompanied with a numerical example. Moreover, we bring out the error estimation and discuss the limiting case. This revised version was published online in June 2006 with corrections to the Cover Date.     

Block Based Bivariate Blending Rational Interpolation via Symmetric Branched Continued Fractions

This paper constructs a new kind of block based bivariate blending rational interpolation via symmetric branched continued fractions. The construction process may be outlined as follows. The first step is to divide the original set of support points into some subsets (blocks). Then construct each block by using symmetric branched continued fraction. Finally assemble these blocks by Newton’s method to shape the whole interpolation scheme. Our new method offers many flexible bivariate blending rational interpolation schemes which include the classical bivariate Newton’s polynomial interpolation and symmetric branched continued fraction interpolation as its special cases. The block based bivariate blending rational interpolation is in fact a kind of tradeoff between the purely linear interpolation and the purely nonlinear interpolation. Finally, numerical examples are given to show the effectiveness of the proposed method.     

COMPUTATION OF VECTOR VALUED BLENDING RATIONAL INTERPOLANTS

As we know, Newton's interpolation polynomial is based on divided differences which can be calculated recursively by the divided-difference scheme while Thiele 's interpolating continued fractions are geared towards determining a rational function which can also be calculated recursively by so-called inverse differences. In this paper, both Newton's interpolation polynomial and Thiele's interpolating continued fractions are incorporated to yield a kind of bivariate vector valued blending rational interpolants by means of the Samelson inverse. Blending differences are introduced to calculate the blending rational interpolants recursively, algorithm and matrix-valued case are discussed and a numerical example is given to illustrate the efficiency of the algorithm.     

A multivariate QD-like algorithm

The problem of constructing a univariate rational interpolant or Padé approximant for given data can be solved in various equivalent ways: one can compute the explicit solution of the system of interpolation or approximation conditions, or one can start a recursive algorithm, or one can obtain the rational function as the convergent of an interpolating or corresponding continued fraction.In case of multivariate functions general order systems of interpolation conditions for a multivariate rational interpolant and general order systems of approximation conditions for a multivariate Padé approximant were respectively solved in [6] and [9]. Equivalent recursive computation schemes were given in [3] for the rational interpolation case and in [5] for the Padé approximation case. At that moment we stated that the next step was to write the general order rational interpolants and Padé approximants as the convergent of a multivariate continued fraction so that the univariate equivalence of the three main defining techniques was also established for the multivariate case: algebraic relations, recurrence relations, continued fractions. In this paper a multivariate qd-like algorithm is developed that serves this purpose.     

On branched continued fractions rational interpolation over pyramid-typed grids

In this paper, three-term recurrence relations for branched continued fractions are determined. Based on the algorithm of partial inverse differences in tensor-product-like manner, the finite branched continued fractions can be applied to rational interpolation over pyramid-typed grids in R ³. By means of the three-term recurrence relations, a characterization theorem is valid. Then an error estimation is worked out. Based on the relationship between the partial inverse differences and partial reciprocal ones, and the partial reciprocal derivatives as well, the BCFs osculatory interpolation with its algorithm is stated which shows it feasibility of partial derivable functions in BCFs expansion at one point.     

A new algorithm for osculatory rational interpolation

Summary A new algorithm is derived for computing continued fractions whose convergents form the elements of the osculatory rational interpolation table.     

QD-Type algorithms for the nonnormal Newton-Padé approximation table

It is well known that solutions of the rational interpolation problem or Newton-Padé approximation problem can be represented with the help of continued fractions if certain normality assumptions are satisfied. By comparing two interpolating continued fractions, one obtains a recursive QD-type scheme for computing the required coefficients. In this paper a uniform approach is given for two different interpolating continued fractions of ascending and descending type, generalizing ideas of Rutishauser, Gragg, Claessens, and others. In the nonnormal case some of the interpolants are equal yielding so-called singular blocks. By appropriate “skips” in the Newton-Padé table modified interpolating continued fractions are derived which involve polynomials known from the Kronecker algorithm and from the Werner-Gutknecht algorithm as well as from the modification of the cross-rule proposed recently by the authors. A corresponding QD-type algorithm for the nonnormal Newton-Padé table is presented. Finally, the particular case of Padé approximation is discussed where—as in Cordellier's modified cross-rule—the given recurrence relations become simpler.     

An algorithm of infinite sums representations and Tasoev continued fractions

For any given real number, its corresponding continued fraction is unique. However, given an arbitrary continued fraction, there has been no general way to identify its corresponding real number. In this paper we shall show a general algorithm from continued fractions to real numbers via infinite sums representations. Using this algorithm, we obtain some new Tasoev continued fractions.