Modeling in Physical Sciences Problems by Single Python Program

Abstract

A Python program was first developed for multiple regression, but it is equally good for modeling of physical sciences phenomena through linear regression, non-linear, and polynomial fitting. The program was implemented on three problem from three disciplines. An example from environmental science is used to implement multiple regression; relative humidity was predicted from daily temperatures and Dew point. The polynomial fitting example was taken from Physics, in which a quadratic polynomial was fitted. The non-linear fitting example was taken from Chemistry, in which the reaction rate is a non-linear function of the product's concentration. Good results have been obtained in each case. The program can do multiple tasks, can easily be understood, and can easily be operated. Another important feature of the program is that it calculates the coefficients through a single line code.