Library of mathematical functions for the language of data-driven functional parallel programs Pythagor

Бесплатный доступ

The language of data-driven functional parallel programs Pythagor is an original programming language, which key features are the absence of variables and loop operators, support for operation-level parallelism, operating by availability. The Pythagor language tools are developed, and a repository of functions is worked out. A library of mathematical functions in terms of functionality comparable to the library math.h of C language has been created. Function calculation is based on the Maclaurin series and reduction formulas. Some mathematical functions are implemented in two versions, one of which performs fast and less accurate calculations, and the other involves accurate and slower calculations. The implementation of the library of mathematical functions for the Pythagor language has been performed for the first time; the calculation of mathematical functions supports parallelization capabilities at the level of operations. Mathematical functions are included in the open repository of the Pythagor language.

Еще

Functional programming, streaming programming, parallel programming, architecture-independent programming, operation-level parallelism, algorithms of mathematical functions, approximate calculations, maclaurin series, function repository, data control flow graph

Еще

Короткий адрес: https://sciup.org/148308951

IDR: 148308951   |   DOI: 10.18101/2304-5728-2019-4-57-64

Статья научная