Using Similarity Metrics to Quantify Differences in High-Throughput Data Sets: Application to X-ray Diffraction Patterns.

The objective of this research is to demonstrate how similarity metrics can be used to quantify differences between sets of diffraction patterns. A set of 49 similarity metrics is implemented to analyze and quantify similarities between different Gaussian-based peak responses, as a surrogate for different characteristics in X-ray diffraction (XRD) patterns. A methodological approach was used to identify and demonstrate how sensitive these metrics are to expected peak features. By performing hierarchical clustering analysis, it is shown that most behaviors lead to unrelated metric responses. For instance, the results show that the Clark metric is consistently one of the most sensitive metrics to synthetic single peak changes. Furthermore, as an example of its utility, a framework is outlined for analyzing structural changes because of size convergence and isotropic straining, as calculated through the virtual XRD patterns.