File: test_verlet_thermostats_asap.py

package info (click to toggle)
python-ase 3.21.1-2
  • links: PTS, VCS
  • area: main
  • in suites: bullseye
  • size: 13,936 kB
  • sloc: python: 122,428; xml: 946; makefile: 111; javascript: 47
file content (89 lines) | stat: -rw-r--r-- 3,360 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
import numpy as np
from ase.build import bulk
from ase.units import fs
from ase.md import VelocityVerlet
from ase.md import Langevin
from ase.md import Andersen
from ase.io import Trajectory, read
from ase.utils import seterr
import pytest
from ase.md.velocitydistribution import (MaxwellBoltzmannDistribution,
                                         Stationary)


# test Verlet, Langevin and Andersen with asap3
@pytest.mark.slow
def test_verlet_thermostats_asap(asap3):
    rng = np.random.RandomState(0)
    calculator = asap3.EMT()
    T_low = 10
    T_high = 300
    md_kwargs = {'timestep': 0.5 * fs, 'logfile': '-', 'loginterval': 500}
    with seterr(all='raise'):
        a = bulk('Au').repeat((4, 4, 4))
        a[5].symbol = 'Ag'

        # test thermalization by MaxwellBoltzmannDistribution
        thermalize(T_high, a, rng)
        assert abs(a.get_temperature() - T_high) < 0.0001

        # test conservation of total energy e0 using Verlet
        a_verlet, traj = prepare_md(a, calculator)
        e0 = a_verlet.get_total_energy()
        md = VelocityVerlet(a_verlet, **md_kwargs)
        md.attach(traj.write, 100)
        md.run(steps=10000)
        traj_verlet = read('Au7Ag.traj', index=':')
        assert abs(traj_verlet[-1].get_total_energy() - e0) < 0.0001

        # test reproduction of Verlet by Langevin and Andersen for thermostats
        # switched off
        pos_verlet = [t.get_positions() for t in traj_verlet[:3]]
        md_kwargs.update({'temperature_K': T_high})
        for MDalgo in [Langevin, Andersen]:
            a_md, traj = prepare_md(a, calculator)
            if MDalgo is Langevin:
                md = MDalgo(a_md, friction=0.0, rng=rng, **md_kwargs)
            elif MDalgo is Andersen:
                md = MDalgo(a_md, andersen_prob=0.0, rng=rng, **md_kwargs)
            md.attach(traj, 100)
            md.run(steps=200)
            traj_md = read('Au7Ag.traj', index=':')
            pos_md = [t.get_positions() for t in traj_md[:3]]
            assert np.allclose(pos_verlet, pos_md)  # Verlet reproduced?

            # test thermalization to target temperature by thermostats and
            # conservation of average temperature by thermostats
            md.set_timestep(4 * fs)
            if MDalgo is Langevin:
                md.set_friction(0.01)
            elif MDalgo is Andersen:
                md.set_andersen_prob(0.01)
            thermalize(T_low, a_md, rng)  # thermalize with low temperature (T)
            assert abs(a_md.get_temperature() - T_low) < 0.0001

            md.run(steps=500)  # equilibration, i.e. thermalization to high T
            temp = []

            def recorder():
                temp.append(a_md.get_temperature())
            md.attach(recorder, interval=1)
            md.run(7000)
            temp = np.array(temp)
            avgtemp = np.mean(temp)
            fluct = np.std(temp)
            print("Temperature is {:.2f} K +/- {:.2f} K".format(avgtemp, fluct))
            assert abs(avgtemp - T_high) < 10.0


def prepare_md(atoms, calculator):
    a_md = atoms.copy()
    a_md.calc = calculator
    traj = Trajectory('Au7Ag.traj', 'w', a_md)
    return a_md, traj


def thermalize(temp, atoms, rng):
    MaxwellBoltzmannDistribution(atoms, temperature_K=temp, force_temp=True,
                                 rng=rng)
    Stationary(atoms)