Adapting the Poisson-Influenced K-Means Algorithm for a Larger User Base
Brian P. Morris, Zachary H. Levine
A superconducting transition edge sensor (TES) can be a useful tool for counting the number of photons in a highly attenuated pulse of light, but it requires calibration for its outputs to be interpretable as photon numbers. The Poisson-Influenced K-Means Algorithm (PIKA) was created to calibrate a TES given an amount of information fundamentally limited by the nature of the device. We present a new implementation of PIKA in a Mathematica notebook, with integrated documentation and explanation, a new system for handling input and output, and corrections to bugs and errors that were present in the original version. This version of the algorithm is designed to be more transparent, accessible, and adaptable than its predecessor. The goal is enhanced ease of use and modification for researchers needing the original functionality of TES calibration, for those working on related applications, and for those working on unrelated applications to which the new implementation of PIKA can be adapted.