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
|
import pickle
import matplotlib.pyplot as plt
import numpy as np
with open('synapse_creation_times_brian1.pickle', 'r') as f:
brian1_results = pickle.load(f)
with open('synapse_creation_times_brian2.pickle', 'r') as f:
brian2_results = pickle.load(f)
results = brian1_results
results.update(brian2_results)
conditions = [('Full', True),
('Full (no-self)', 'i != j'),
('One-to-one', 'i == j'),
('Simple neighbourhood', 'abs(i-j) < 5'),
('Gauss neighbourhood', 'exp(-(i - j)**2/5) > 0.005'),
('Random (50%)', 0.5),
('Random (10%)', 0.1),
('Random (1%)', 0.01),
('Random no-self (50%)', '(i != j) * 0.5'),
('Random no-self (10%)', '(i != j) * 0.1'),
('Random no-self (1%)', '(i != j) * 0.01')]
targets = ['PythonLanguage', 'CPPLanguage', 'Brian 1']
# nicer names for the labels
lang_translation = {'PythonLanguage': 'Brian 2 (Python)',
'CPPLanguage': 'Brian 2 (C++)',
'Brian 1': 'Brian 1'}
# Do some plots
for pattern, condition in conditions:
plt.figure()
for lang_name in targets:
data = [(connections, time) for ((lang, connections, p), time) in results.items()
if lang == lang_name and p == pattern]
data.sort(key=lambda item: item[0])
data = np.array(data).T
plt.plot(data[0], data[1], 'o-', label=lang_translation[lang_name])
plt.xscale('log')
plt.yscale('log')
plt.xlabel('Number of created connections')
plt.ylabel('time (s)')
plt.legend(loc='best', frameon=False)
plt.title('%s: "%s"' % (pattern, condition))
plt.savefig('plots/%s.png' % pattern.replace(' ', '_'))
|
