2.6. Statistical Data Analysis

For the classification of olive oil samples according to their geographical origin (Tunisia, Spain, and France), a multivariate analysis was performed using SIMCA V16.0.1. The Principal Components Analysis (PCA) is an unsupervised method aiming to find new variables called dimensions calculated from a covariance matrix of the original variables while preserving as much as possible the statistical information (variability), allowing summary and visualization of the information in large datasets. The samples classification was accomplished by the PCA score plot, and the determination of the most discriminating elements was conducted by the loadings plot. Spearman’s correlation was calculated for the trace element concentrations in an exchangeable fraction of soil and corresponding olive oils in order to verify the possible correlation between both. Spearman’s correlation factors were calculated using OriginLab 2018.