Summation of Power Series by Self-Similar Factor Approximants

TL;DR

This paper introduces a new summation method for power series using self-similar factor approximants, enabling accurate extrapolation from small to large variable domains in physics-related problems.

Contribution

It develops a novel summation technique based on self-similar approximation theory, transforming series into products for better extrapolation.

Findings

Effective in statistical mechanics and condensed matter problems

Accurately extrapolates series from small to large variables

Applicable to many-body theory series

Abstract

A novel method of summation for power series is developed. The method is based on the self-similar approximation theory. The trick employed is in transforming, first, a series expansion into a product expansion and in applying the self-similar renormalization to the latter rather to the former. This results in self-similar factor approximants extrapolating the sought functions from the region of asymptotically small variables to their whole domains. The method of constructing crossover formulas, interpolating between small and large values of variables is also analysed. The techniques are illustrated on different series which are typical of problems in statistical mechanics, condensed-matter physics, and, generally, in many-body theory.