A Beer’s Law Problem Integrating Multiple Linear Regression and Simplex Optimization

TCE

The Chemical Educator

ISSN: 1430-4171 (electronic version)

Table of Contents

Abstract Volume 20 (2015) pp 113-116

A Beer’s Law Problem Integrating Multiple Linear Regression and Simplex Optimization

Warren Wm. McGovern and Eugene T. Smith*

Harriet L. Wilkes Honors College, Florida Atlantic University, Jupiter, FL 33458, esmith@fau.edu
Received January 22, 2015. Accepted April 13, 2015.

Published: 1 May 2015

Abstract. Simultaneously obtaining specie concentrations in a mixture is conveniently determined spectroscopically from Beer’s law using multiple linear regressions. Unfortunately, numerical methods can produce answers that don’t make physical sense, i.e. negative concentrations. In this investigation, simplex optimization is employed, in conjunction with multiple linear regressions, to find alternative minimized answers that do not contain negative values. An Excel macro-containing workbook is supplied which contains coding for simplex optimization.

Key Words: Computers in Chemistry; analytical chemistry

(*) Corresponding author. (E-mail: esmith@fau.edu )

Article in PDF format (188 KB) HTML format

Supporting Materials:

Two Excel files are available; Workbook for simplex optimization, and Documentation for simplex optimization workbook. A text file containing the VBA source code is included. (107 KB)