Example jupyter notebooks detailing the simulation and reconstruction of synthetic noisy time series trajectories from molecular dynamics training data.
Run in order:
- 0: extract head to tail distances and time averaged trajectories ((a) to (b)).
- 1: Learn low dimensional representations of these trajectories ((a) to (c)).
- 3: Learn low dimensional representations of dynamics from noisy time averaged head to tail data ((b) to (d)).
- 4: Learn intermanifold map between manifolds learned in 1 and 3 ((d) to (c)).
- 5: Learn backmaping between reconstructed manifold learned in 4 and all atom positions ((c) to (e)).