forked from pmla/polyhedral-template-matching
/
ptm_benchmark.py
179 lines (135 loc) · 4.84 KB
/
ptm_benchmark.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
import ptmmodule
import numpy as np
import struct
import scipy.linalg
import matplotlib.pyplot as plt
import scipy.optimize
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
from mpl_toolkits.mplot3d.art3d import Poly3DCollection
def plot_points(points, label=False, edges=None):
fig = plt.figure()#figsize=(16,14))
fig.set_tight_layout(True)
ax = fig.add_subplot(111, projection='3d')
(xs, ys, zs) = zip(*points)
ax.scatter(xs, ys, zs, c='r')
if edges:
for (a, b) in edges:
p = [points[a], points[b]]
(xs, ys, zs) = zip(*p)
plt.plot(xs, ys, zs, c='k')
if label:
for p, l in zip(points, label):
ax.text(p[0], p[1], p[2], l, size=30, color='k')
(xs, ys, zs) = zip(*points)
lim = max([abs(e) for e in xs+ys+zs])
ax.set_xlim(-lim, lim)
ax.set_ylim(-lim, lim)
ax.set_zlim(-lim, lim)
plt.show()
def vonmises(P):
t = (P[0][0] - P[1][1])**2 + (P[1][1] - P[2][2])**2 + (P[2][2] - P[0][0])**2
t += 6 * (P[1][2]**2 + P[2][0]**2 + P[0][1]**2)
return np.sqrt(t / 2)
def calc_fcc_strain(cF, cF_res):
F = np.array(cF).reshape((3, 3))
r = np.array(cF_res)
(U, A) = scipy.linalg.polar(F, side='left')
dot = np.dot(np.cross(U[0], U[1]), U[2])
if dot < 0:
raise Exception("nope")
return (vonmises(A), sum(r))
def calc_fcc_strain(cF, cF_res, cP, cU, neighbours, mapping):
points_fcc = [ [ 0. , 0.766032346285, 0.766032346285],
[ 0. , -0.766032346285, -0.766032346285],
[ 0. , 0.766032346285, -0.766032346285],
[ 0. , -0.766032346285, 0.766032346285],
[ 0.766032346285, 0. , 0.766032346285],
[ -0.766032346285, 0. , -0.766032346285],
[ 0.766032346285, 0. , -0.766032346285],
[ -0.766032346285, 0. , 0.766032346285],
[ 0.766032346285, 0.766032346285, 0. ],
[ -0.766032346285, -0.766032346285, 0. ],
[ 0.766032346285, -0.766032346285, 0. ],
[ -0.766032346285, 0.766032346285, 0. ],
[ 0. , 0. , 0. ] ]
points_fcc = np.array(points_fcc)
neighbours = np.concatenate((neighbours[1:], neighbours[:1]))
neighbours -= np.mean(neighbours, axis=0)
scale = np.mean(np.linalg.norm(neighbours, axis=1))
for i in range(len(neighbours)):
neighbours[i] /= scale
neighbours = neighbours[list(mapping)]
res = np.linalg.lstsq(points_fcc, neighbours)
_F = res[0].T
F = np.array(cF).reshape((3, 3))
r = np.array(cF_res)
(U, P) = scipy.linalg.polar(F, side='left')
dot = np.dot(np.cross(U[0], U[1]), U[2])
if dot < 0:
raise Exception("nope")
assert(np.linalg.norm(F - _F) < 1E-6)
cU = np.array(cU).reshape((3, 3))
cP = np.array(cP).reshape((3, 3))
print cP - P
print np.linalg.norm(U, axis=1)
return (vonmises(P), sum(r))
def calc_fcc_strain(cF, cF_res, cP, cU, neighbours):
F = np.array(cF).reshape((3, 3))
r = np.array(cF_res)
(U, P) = scipy.linalg.polar(F, side='left')
dot = np.dot(np.cross(U[0], U[1]), U[2])
if dot < 0:
raise Exception("nope")
U = np.array(cU).reshape((3, 3))
P = np.array(cP).reshape((3, 3))
return (vonmises(P), sum(r))
def run(pos, nbrs):
num_atoms = len(pos)
result = np.zeros(num_atoms, int)
strains = np.zeros(num_atoms).astype(np.double)
rmsds = np.zeros(num_atoms).astype(np.double)
if 1:
for i in range(num_atoms):
positions = np.concatenate(([pos[i]], pos[nbrs[i][:18]]))
(struct, alloy, rmsd, scale, rot, F, F_res, P, U, lattice_constant) = ptmmodule.index_structure(positions, calculate_strains=1, topological_ordering=1)
if 0 and struct == 0 and all([e > 1000 for e in nbrs[i]]):
print nbrs[i]
print positions
positions -= pos[i]
plot_points(positions[:15])
#(struct, alloy, rmsd, scale, rot, lattice_constant) = ptmmodule.index_structure(positions)
#if struct == 2:
# vm, r = calc_fcc_strain(F, F_res, P, U, positions[:13])
# strains[i] = vm
result[i] = struct
rmsds[i] = rmsd
indices = np.where(rmsds < 0.12)[0]
kept = np.bincount(result[indices])
kept[0] += num_atoms - len(indices)
print "rmsd < 0.12:", kept, sum(kept)
return result, rmsds
indices = np.where(result == 2)[0]
plt.hist(strains[indices], bins=200)
plt.show()
def go():
if 1:
dat_pos = open('test_data/FeCu_positions.dat', 'rb').read()
dat_nbr = open('test_data/FeCu_nbrs.dat', 'rb').read()
#dat_pos = open('test_data/fcc_positions.dat', 'rb').read()
#dat_nbr = open('test_data/fcc_nbrs.dat', 'rb').read()
n = len(dat_pos) / 24
print "num atoms:", n
pos = np.array(struct.unpack(n * 3 * "d", dat_pos)).reshape((n, 3))
nbrs = np.array(struct.unpack(n * 24 * "i", dat_nbr)).reshape((n, 24))
n = len(dat_pos) / 24
print "num atoms:", n
#pos = np.load('positions0.npy')
#nbrs = np.load('neighbours0.npy')
ptm, rmsds = run(pos, nbrs)
print ptm
print np.bincount(ptm)
np.save('structures', ptm)
np.save('rmsds', rmsds)
if __name__ == "__main__":
go()