On ``pre-historic linear programming and the figure of the Earth

DOI10.1007/S10957-017-1165-5MaRDI QIDQ1682983zbMATH OpenOpenAlexFDO

Publication date 6 December 2017

Published in Journal of Optimization Theory and Applications (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1007/s10957-017-1165-5

geodesy Laplace least squares minimizing errors regression line shape of the Earth

History of mathematics in the 18th century (01A50) History of statistics (62-03) Linear regression; mixed models (62J05) History of operations research and mathematical programming (90-03) Linear programming (90C05)

Review text

The author explores a well-studied topic of Roger Boscovich's line of best fit defined by minimizing the sum of the absolute values of deviations of measurements. Boscovich invented this method while examining meridian measurements related to attempts to determine the exact shape of the Earth. The contribution of the present paper is in that it gives a detailed mathematical presentation of the method and its modifications (by Laplace for instance), adjusted to today's mathematical lingo; it also demonstrates how linear programming methods can be used to treat the Boscovich constrained minimization problem of errors. In addition, the author poses two fair questions: 1. What was the first publication to mention the simplex method of linear programing as a solution method for problems of the sort? 2. Which of the algorithms published before, say, 1951, have been -- or can be -- interpreted in terms of the simplex method (or some variant thereof)? He proposes that the first to mention linear programming was T. E. Harris in 1950 and opines that the second question requires further investigation, while citing several other sources that are in the domain of this question. \textit{V. Varićak} [``Matematički rad Boškovićev, Rad JAZU 181, 75--208 (1910)] writes about these measurements of Boscovich and Maire in detail including reactions of Boscovich's critics Zach, Ricchebach and Secchi. Most of the literature cited in the paper under review and its purposes are shared with, for instance, a monograph by \textit{Ž. Marković} [Rudje Bošković (Serbo-Croat). Zagreb: JAZU (1968)]. The reviewer presented a talk titled ``R. J. Boscovich and the line of best fit at the Joint AMS/MAA Meeting, Washington D.C., January 2000, with the similar content as the paper under review. It was argued that the functions \(\sum |x-x_i|\), in most cases lend themselves to constructs in linear programming that are in turn at the core of this method of Boscovich (this connection is also mentioned, en passant, in the reviewer's short monograph [Rudjer Bošković (Serbian). Pittsburgh, PA: Helios Publishing Co. (2006; Zbl 1160.01017)]. The reviewer notes that one should also look at [\textit{N. Abakumov}, ``Astronomsko-geodetski radovi Rudjera Boškovića, Almanah Bošković, 192--199 (1950)] in relation to question 1. As to question 2, clearly, new disciplines and new methods often offer alternative or easier solutions to problems posed in the yesteryear. One has, however, to guard against a natural temptation of clothing the old ideas, methods and theories in new garbs, without regard to the existing conditions of science at the time of these original inventions.

Recommendations

Cites work

Cited in

(6)

This page was built for publication: On ``pre-historic linear programming and the figure of the Earth

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1682983)