Applying Best Practice Processes to Your Curve Fitting and Data Analysis
Extracting the most accurate results from experimental data during curve fitting and subsequent analysis requires an understanding of the quality of the data (factors such as outlying data points, poorly fitting curves and incomplete data sets) and optimizing models before routine fitting analysis begins.
XLfit enables the expert modeler, the student and the novice to experiment with data, define best fits, visualize their fit data and rapidly fit multiple curve models – including secondary fit analysis – without the need to leave the Excel environment.
This first article in our Best Practice series examines ways to increase confidence in the accuracy of curve-fitting methods and describes how to avoid some of the common causes of unsuccessful curve fits.
Explore different ways to solve curve fitting issues, such as using ‘knock out’ of outlying data points to ensure an accurate fit. This article also examines the criteria used to select the correct fit model.
Explore the relative advantages and disadvantages of various fitting methodologies and provides guidance to users on selection of the most suitable fit models.
Understand the importance of parameters and their starting values in the fitting process. The concept of residuals, the criteria used to determine ‘goodness of fit’, is also introduced.
Learn why robust fitting is becoming a popular method of fitting data while neutralizing the effect of outliers in a data set.
Find out how global fitting combines multiple fits to find the optimum global parameter value for each of the fits.