<p>In earth sciences, measurements usually produce compositional data with a property called closedness. Researchers who use common statistical methods on compositional data ignore spurious correlations, which causes incorrect results. This article introduces transforma More
<p>In earth sciences, measurements usually produce compositional data with a property called closedness. Researchers who use common statistical methods on compositional data ignore spurious correlations, which causes incorrect results. This article introduces transformations for opening closed system of compositional data. These transformations include the additive logarithmic ratio (alr), the centred logarithmic ratio (clr), and the isometric logarithmic ratio (ilr). They are all defined in terms of logarithms of ratios. We then applied the clr transformation to a soil chemical data set. We also analysed the results of applying cluster analysis on the clr transformed data using Spearman's correlation coefficient matrix as distance. We also investigated how applying clr transformation affects spurious correlation, skewness and outliers in the data using R statistical software.</p>
Manuscript profile
