SyncPy

Abstract

Despite the growing interest on synchrony in many scientific fields, at the present, a unified framework and tool to study synchrony are still missing. This paper introduces the SyncPy library, an open-source Python library, conceived to perform interpersonal synchrony analysis on time series collected during dyadic and/or multiparty human-human/human-machine interaction. First, we introduce the philosophy of the library; then, we provide a formal description of the main technical aspects the library is grounded on and its current functionalities. Finally, we show concrete examples of synchrony analysis on time series from real experiments involving both human and robotic partners. SyncPy is at its early stage of developement, but it is the very first open-source attempt to collect together knowledge on synchrony. We expect this library will be extended and improved in the near future with the contribution of its community of users.