-
Notifications
You must be signed in to change notification settings - Fork 10
/
visualize.py
180 lines (149 loc) · 7.28 KB
/
visualize.py
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
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
import itertools
import collections
import numpy as np
from sklearn import manifold
import matplotlib.pyplot as plt
import matplotlib.ticker as ticker
import pymatgen as pmg
from .db_actions import filter_evaluations, list_evaluations
def normalize_parameters(optimization_bounds, parameters_array):
optimization_bounds = np.array(optimization_bounds)
min_norm_vector = optimization_bounds[:, 0]
range_norm_vector = optimization_bounds[:, 1] - optimization_bounds[:, 0]
return (parameters_array - min_norm_vector) / range_norm_vector
def tsne_evaluations(dbm, ax, potential, run_id, limit=1000, condition='best', labels=None):
cursor = filter_evaluations(dbm, potential, limit, condition, run_id, labels)
parameters_array = []
scores = []
for row in cursor:
parameters_array.append(row['parameters'])
scores.append(row['score'])
parameters_array = np.array(parameters_array)
scores = np.array(scores)
with dbm.connection:
optimization_bounds = dbm.connection.execute(
'SELECT bounds FROM run WHERE id = ?', (run_id,)).fetchone()['bounds']
normalized_parameters = normalize_parameters(optimization_bounds, parameters_array)
tsne = manifold.TSNE(n_components=2, init='pca', random_state=0)
Y = tsne.fit_transform(normalized_parameters)
return ax.scatter(Y[:, 0], Y[:, 1], c=scores, cmap=plt.cm.Spectral, alpha=0.2)
def visualize_progress(dbm, run_id, window=100, title=None, filename=None, show=True):
df = list_evaluations(dbm, run_id)
title = title or f'Optimization progress for run {run_id}'
linewidth = 0.5
fig, axes = plt.subplots(1, 3, sharey=True, sharex=True)
axes[0].plot(df['score'].rolling(window=window, center=False).mean(), linewidth=linewidth)
axes[0].set_title('mean')
axes[0].set_ylabel('normalized score')
axes[0].set_xlabel('iterations')
axes[0].ticklabel_format(axis='x', style='sci', scilimits=(-2, 2))
axes[1].plot(df['score'].rolling(window=window, center=False).median(), linewidth=linewidth)
axes[1].set_title('median')
axes[1].set_xlabel('iterations')
axes[2].plot(df['score'].rolling(window=window, center=False).min(), linewidth=linewidth)
axes[2].set_title('minimum')
axes[2].set_xlabel('iterations')
fig.set_size_inches((15, 5))
if filename:
fig.savefig(filename, transparent=True)
if show:
plt.show()
def visualize_single_calculation(dft_calculations, md_calculations, plot='forces', show=True, filename=None):
dft_forces = []
md_forces = []
dft_c = []
md_c = []
c_ij = [(0, 0), (1, 1), (2, 2), (0, 1), (0, 2), (1, 2)]
for dft_calc, md_calc in zip(dft_calculations, md_calculations):
dft_forces.extend(np.linalg.norm(dft_calc.forces, axis=1).tolist())
md_forces.extend(np.linalg.norm(md_calc.forces, axis=1).tolist())
dft_c.append([dft_calc.stress[ij] for ij in c_ij])
md_c.append([md_calc.stress[ij] for ij in c_ij])
dft_rel_energies = []
md_rel_energies = []
for (dft_calc_i, md_calc_i), (dft_calc_j, md_calc_j) in itertools.combinations(zip(dft_calculations, md_calculations), 2):
dft_rel_energies.append(dft_calc_i.energy - dft_calc_j.energy)
md_rel_energies.append(md_calc_i.energy - md_calc_j.energy)
point_alpha = 0.1
point_size = 3
def plot_equal_line(axes, x, y):
axes.plot([np.min(x), np.max(x)], [np.min(x), np.max(x)], '--', color='#E65933', alpha=0.5)
if plot == 'forces':
fig, axes = plt.subplots(1, 1)
axes.scatter(dft_forces, md_forces, alpha=point_alpha, s=point_size)
axes.set_xlabel(r'DFT force magnitude [$eV/\AA$]')
axes.set_ylabel(r'MD force magnitude [$eV/\AA$]')
plot_equal_line(axes, dft_forces, md_forces)
elif plot == 'energy':
fig, axes = plt.subplots(1, 1)
axes.scatter(dft_rel_energies, md_rel_energies, alpha=point_alpha, s=point_size)
axes.set_xlabel(r'DFT relative energies [$eV$]')
axes.set_ylabel(r'MD relative energies [$eV$]')
plot_equal_line(axes, dft_rel_energies, md_rel_energies)
elif plot == 'stress':
fig, axes = plt.subplots(2, 3)
axes[0,0].scatter(dft_c[0], md_c[0])
axes[0,0].set_xlabel(r'DFT $\sigma_{xx} [bar]$')
axes[0,0].set_ylabel(r'MD $\sigma_{xx} [bar]$')
plot_equal_line(axes[0,0], dft_c[0], md_c[0])
axes[0,1].scatter(dft_c[1], md_c[1])
axes[0,1].set_xlabel(r'DFT $\sigma_{yy} [bar]$')
axes[0,1].set_ylabel(r'MD $\sigma_{yy} [bar]$')
plot_equal_line(axes[0,1], dft_c[1], md_c[1])
axes[0,2].scatter(dft_c[2], md_c[2])
axes[0,2].set_xlabel(r'DFT $\sigma_{zz} [bar]$')
axes[0,2].set_ylabel(r'MD $\sigma_{zz} [bar]$')
plot_equal_line(axes[0,2], dft_c[2], md_c[2])
axes[1,0].scatter(dft_c[3], md_c[3])
axes[1,0].set_xlabel(r'DFT $\sigma_{xy} [bar]$')
axes[1,0].set_ylabel(r'MD $\sigma_{xy} [bar]$')
plot_equal_line(axes[1,0], dft_c[3], md_c[3])
axes[1,1].scatter(dft_c[4], md_c[4])
axes[1,1].set_xlabel(r'DFT $\sigma_{xz} [bar]$')
axes[1,1].set_ylabel(r'MD $\sigma_{xz} [bar]$')
plot_equal_line(axes[1,1], dft_c[4], md_c[4])
axes[1,2].scatter(dft_c[5], md_c[5])
axes[1,2].set_xlabel(r'DFT $\sigma_{yz} [bar]$')
axes[1,2].set_ylabel(r'MD $\sigma_{yz} [bar]$')
plot_equal_line(axes[1,2], dft_c[5], md_c[5])
fig.set_size_inches((12, 15))
plt.tight_layout()
if filename:
fig.savefig(filename, transparen=True)
if show:
plt.show()
return fig, axes
def visualize_radial_pair_distribution(calculations, distance=10, filename=None, show=True):
"""visualize the radial pair distributions for each atom in training set"""
specie_types = set()
for calculation in calculations:
specie_types.update(calculation.structure.types_of_specie)
distances = collections.defaultdict(list)
for calculation in calculations:
for original_site, sites in zip(calculation.structure, calculation.structure.get_all_neighbors(10, include_index=True)):
for (site, distance, index) in sites:
distances[tuple(sorted([original_site.specie, site.specie]))].append(distance)
fig, axes = plt.subplots(len(distances), 1, sharex=True, sharey=True)
for pair, ax in zip(distances, axes.ravel()):
ax.set_xlabel('distance [angstroms]')
ax.set_ylabel('count')
ax.set_title(f'{pair[0].symbol} - {pair[1].symbol}')
ax.hist(distances[pair], bins=int(distance*10))
fig.set_size_inches((10, len(distances)*4))
if filename:
fig.savefig(filename, transparent=True)
if show:
plt.show()
return fig, axes
def visualize_pair_energies(separations, pair_energies, filename=None, show=True):
fig, axes = plt.subplots(len(pair_energies), 1, sharex=True, sharey=True)
for ax, (label, energies) in zip(np.ravel(axes), pair_energies.items()):
ax.plot(separations, energies)
ax.axhline(y=0, c='k', linewidth=1, linestyle='dotted')
ax.set_ylabel('Energy [eV]')
ax.set_yscale('symlog')
while len(ax.get_yticks()) > 5: # would be nice if i could just specify 5...
ax.set_yticks(ax.get_yticks()[::2])
ax.set_title(label)
axes[-1].set_xlabel(r'Separation [$\AA$]')
plt.show()