{"title":"FastAAA: A fast rational-function fitter","authors":"A. Hochman","doi":"10.1109/EPEPS.2017.8329756","DOIUrl":null,"url":null,"abstract":"FastAAA is an algorithm for fitting rational-functions to a set of N data samples. In each step of the algorithm, it computes an order n fit via a fast, O(Nn), update of the previous order n − 1 fit. The algorithm stops at the first order that yields an acceptable error. The errors of the fits of orders 1… n, are evaluated in O(Nn2) operations. If the data can be represented exactly with n poles, the algorithm is guaranteed to stop after n iterations (in exact arithmetic). It is possible to fit p rational-functions, sharing the same set of poles, to p sets of data, in O(pNn2) operations. The stability of the poles and Hermitian symmetry of the fit can be guaranteed.","PeriodicalId":397179,"journal":{"name":"2017 IEEE 26th Conference on Electrical Performance of Electronic Packaging and Systems (EPEPS)","volume":"98 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2017-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"12","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2017 IEEE 26th Conference on Electrical Performance of Electronic Packaging and Systems (EPEPS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/EPEPS.2017.8329756","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 12
Abstract
FastAAA is an algorithm for fitting rational-functions to a set of N data samples. In each step of the algorithm, it computes an order n fit via a fast, O(Nn), update of the previous order n − 1 fit. The algorithm stops at the first order that yields an acceptable error. The errors of the fits of orders 1… n, are evaluated in O(Nn2) operations. If the data can be represented exactly with n poles, the algorithm is guaranteed to stop after n iterations (in exact arithmetic). It is possible to fit p rational-functions, sharing the same set of poles, to p sets of data, in O(pNn2) operations. The stability of the poles and Hermitian symmetry of the fit can be guaranteed.