SigCorr: A Python package for studies of trials factors

SigCorr is a framework that implements a wide range of tools to conduct studies of the trials factor, which is an important parameter in the statistical characterization of searches for new phenonema, such as a hypothetical elementary particle or astrophysical source. Such searches are commonly carried out for a range of possible signal hypotheses. The statistical significance of the most significant observation (generally a localized excess of data in a broad background distribution) must take into account the possibility that the background could have fluctuated anywhere in the search region, not just at the point of maximum interest. Taking into account this so-called «look-elsewhere effect» is performed by correcting the corresponding p-value from its local to a global value. The ratio of the former to the latter is known in the statistical literature as the trials factor.