Are LIGO data connected?

The analysis of ultra-precise data of stellar light variations observed with satellites (CoRoT, Kepler, SoHO, etc.) has revealed some unexpected results that cannot be explained by theory. This hampers and, in some cases, even prevents from interpreting the stellar oscillations. We have tackled this problem from the analysis side, testing the mathematical conditions necessary to properly apply the harmonic analysis techniques to the data. In particular we have developed and algorithm to calculate what we have called “connectivities” for testing the analyticity of the underlying function from which the time series are a discrete sample. For the cases studied the result of the analyticity test is negative providing a counterexample of the overextended paradigm based on the assumption that any sequence of measures can be interpreted as the Fourier expansion of an analytic function. In this work we present the results of the analyticity test performed to the data of the first detection of gravitational waves in which, the data analysis, as delivered by the Advanced LIGO instrument team, relies upon the previously cited paradigm. We found that both raw and filtered data are not connected, i.e. the function underlying the sampled data is non-analytic. We discuss here the mathematical and physical consequences of this result.