Tomography of scaling

Scaling describes how a given quantity Y that characterizes a system varies with its size P . For most complex systems, it is of the form Y ∼ P β with a non-trivial value of the exponent β , usually determined by regression methods. The presence of noise can make it difficult to conclude about the existence of a nonlinear behaviour with β ≠ 1 and we propose here to circumvent fitting problems by investigating how two different systems of sizes P 1 and P 2 are related to each other. This leads us to define a local scaling exponent β loc that we study versus the ratio P 2 / P 1 and provides some sort of ‘tomography scan’ of scaling across different values of the size ratio, allowing us to assess the relevance of nonlinearity in the system and to identify an effective exponent that minimizes the error for predicting the value of Y . We illustrate this method on various real-world datasets for cities and show that our method reinforces in some cases the standard analysis, but is also able to provide new insights in inconclusive cases and to detect problems in the scaling form such as the absence of a single scaling exponent or the presence of threshold effects.