Multi-block data analysis using ComDim for the evaluation of complex samples: Characterization of edible oils

Abstract The ComDim chemometrics method for multi-block analysis was employed to evaluate thirty-two vegetable oil samples analyzed by near infrared (NIR) and ultraviolet–visible (UV–Vis) spectroscopy, and by Gas Chromatography with flame ionization detection (GC-FID) for their fatty acids composition. This unsupervised pattern recognition method was able to extract information from the tables of results that could be presented in informative graphs showing the relationship between the samples through the scores, the predominance of information in particular tables through the saliences and the contribution of the variables in each table which were responsible for the similarities observed in the samples, through the loadings plots. It was possible to infer similarities and differences among the samples studied according to the specific absorption in the UV–Vis and NIR region, as well as their fatty acids composition. The proposed methodology demonstrates the applicability of ComDim for the characterization of samples when different variables (different techniques) describe the same samples. In this particular study, the ComDim chemometrics method was able to discriminate samples by their characteristics and compositions.