When the Arc-Cosine transformation is applied, ArrayMiner uses the actual angle between two profile vectors as distance measure, instead of the cosine of the angle between the vectors.
    Both the Pearson Coefficient and the standard correlation use the transformation in ArrayMiner. Indeed, the raw correlations sometimes used in other clustering tools can be shown not to comply with what is called the triangular inequality, which means that such a measure is not a metric in the mathematical sense. The triangular inequality captures the intuitive notion that going from point A to a point B should not be longer than going from A to B via a third point C. Accordingly, clustering results obtained with a non-metric distance measure may be awkward and counterintuitive.

Arc-Cosine Transformation