This is a minimal package for the simulation of coherent phonons. A barebones example goes like so:
import coherentphonons as p
import coherentphonons.splines as psp
import numpy as np
import matplotlib.pyplot as plt
x = np.linspace(-0.1, 0.1, 7)
e = 50 * x ** 2# + 2 * x ** 3
a = psp.make_pes_splines_1d(x, e)
propagator = p.make_propagator_1d(a, 10.0, 2)
sol = p.solve(propagator, (0, 30), [0.01, 0.0])
tfine = np.linspace(0, 30, 1000)
plt.plot(tfine, sol.sol(tfine)[0])
plt.show()TODO:
- Document units. Maybe use pint.
- Support multidimensional potential energy functions/propagators. What is a good way to do this? I currently use radial basis functions in my personal code.
- High dimensional PES training using small neural nets.
- Ensemble excitations for Poincare surfaces of sections.
- More tests. What integration tests? How would that even happen?
- A few examples, both insulating, dispersive and insulating, and dissipative+dispersive.
This is effectively research code and I add features/change API specifications quite often and aggressively. Use with care, reviews appreciated.