Chains method for storage of multidimensional triangulation

Автор: Klyachin Vladimir Aleksandrovich, Popov Vladimir Valentinovich

Журнал: Математическая физика и компьютерное моделирование @mpcm-jvolsu

Рубрика: Прикладная математика

Статья в выпуске: 2 (19), 2013 года.

Бесплатный доступ

This study is an attempt to obtain lower bounds for the amount of memory required to store the data of multi-dimensional triangulations. The author offers a method of organizing storage triangulation, allowing in + 1 times to reduce the amount of memory in comparison with one of the standard methods. The essence of the method proposed by the author is the partitioning of the set of simplices in an array of chains adjacent simplices. Neighborhood relations can be written in a more compact structure of the triangulation. More accurately. If 𝜇(𝑓) denotes the number of bits occupied by storage triangulation points in R𝑛 by one of the standard ways, then lim 𝑁→∞ 𝜇(𝑓) 𝑚log2𝑁 = 𝑛+1, where — number of simplexes. Whereas the proposed method of storage triangulation this limit is 1 for any dimension. To investigate the optimal method, the author formalises the problem and introduces the concept of triangulation as a way to store one-to-one mapping of the set of all triangulations of a set of points in R𝑛 on the set of natural numbers N. Thus, the method of storage — a comparison of each triangulation identify the unique natural number. Study the accuracy of the asymptotic estimates leads to the problem of obtaining lower triangulations given set of points. This article discusses two examples where such a calculation is accurate. In both cases,there is the exponential growth for number of triangulations depended on the number of points. Reference to recent works by foreign mathematicians say about general behavior of the number of such triangulations in the plane. This allows the author to hypothesize the existence of such a storage method of triangulation, in which the memory capacity increases linearly with the number of points. In particular, we expect that the optimal value of the above limit is 0.

Еще

Triangulation, simplex, memory volume estimate, number of triangulations, catalan number

Короткий адрес: https://sciup.org/14968738

IDR: 14968738

Статья научная