Path Weight Sampling implementation for Langevin systems in presence of feedback. Uses Onsager-Machlup formulation for stochastic actions and the Rosenbluth-Rosenbluth method for importance sampling.
Needs Python3, numpy, numba.
To run: nohup python feedbackpws_rr4.py &
Output gets saved as nohup.out. Output is an array where columns correspond to chosen trajectory durations, rows correspond alternately to mutual information and time-integrated transfer entropies, for each of a chosen number of trajectories.