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
|
from collections import defaultdict
from numpy import inf
import ase.db
def analyze(filename, tag='results'):
energies = defaultdict(list)
mintimes = defaultdict(lambda: 999999)
formulas = []
db = ase.db.connect(filename)
for row in db.select(sort='formula'):
if row.formula not in formulas:
formulas.append(row.formula)
energies[row.formula].append(row.get('energy', inf))
emin = {formula: min(energies[formula]) for formula in energies}
data = defaultdict(list)
for row in db.select(sort='formula'):
if row.get('energy', inf) - emin[row.formula] < 0.01:
t = row.t
if row.n < 100:
nsteps = row.n
mintimes[row.formula] = min(mintimes[row.formula], t)
else:
nsteps = 9999
t = inf
else:
nsteps = 9999
t = inf
data[row.optimizer].append((nsteps, t))
print(formulas)
D = sorted(data.items(), key=lambda x: sum(y[0] for y in x[1]))
with open(tag + '-iterations.csv', 'w') as fd:
print('optimizer,' + ','.join(formulas), file=fd)
for o, d in D:
print('{:18},{}'
.format(o, ','.join('{:3}'.format(x[0])
if x[0] < 100 else ' '
for x in d)),
file=fd)
data = {opt: [(n, t / mintimes[f]) for (n, t), f in zip(x, formulas)]
for opt, x in data.items()}
D = sorted(data.items(), key=lambda x: sum(min(y[1], 999) for y in x[1]))
with open(tag + '-time.csv', 'w') as fd:
print('optimizer,' + ','.join(formulas), file=fd)
for o, d in D:
print('{:18},{}'
.format(o, ','.join('{:8.1f}'.format(x[1])
if x[0] < 100 else ' '
for x in d)),
file=fd)
if __name__ == '__main__':
analyze('results.db')
|
