Open-Source Visual Programming Software for Introducing Principal Component Analysis to the Analytical Curriculum

With the increasing complexity of analytical data nowadays, great reliance on statistical and chemometric software is quite common for scientists. Powerful open-source software, such as Python, R, and the commercial software MATLAB, demands good coding skills. Writing original code could be challenging for students with no prior programming experience. Orange Data Mining is a Python based visual programming software that has been used widely in many scientific publications. Principal component analysis (PCA) is one of the most common exploratory data analysis techniques with applications in outlier detection, dimensionality reduction, graphical clustering, and classification. By using a program workflow based on widgets (a computational unit within Orange), the task of PCA can be done very quickly. The same workflow could be used for different types of analytical data without the need for reprogramming again. The application of Orange Data Mining software to PCA exploratory analysis of sugar NIR spectral data from a portable NIR spectrometer will be demonstrated. Further data sets including multivariate coffee composition data, instant coffee FTIR spectra, vegetable oil fatty acid composition, and vegetable oil NMR spectra were given as Supporting Information to enhance the learning of software through repetition. From the demonstration, it can be easily seen how Orange Data Mining software will be useful for introducing PCA to the analytical curriculum.