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
|
import sys
import re
import numpy as np
from collections import defaultdict
groups = defaultdict(list)
for line in open(sys.argv[1]):
gap, chunk, procs, info = re.split("\s+", line, 3)
seconds = re.search("in (.+) seconds", info).groups(0)[0]
if gap == '100' or chunk == '100': continue
groups[(int(gap), int(chunk), int(procs))].append(float(seconds))
bychunk = defaultdict(list)
bygap = defaultdict(list)
bycpu = defaultdict(list)
for gap, chunk, cpu in groups:
if cpu != 4: continue
#if chunk != 5000: continue
m = np.mean(groups[(gap, chunk, cpu)])
groups[(gap, chunk, cpu)] = m
bychunk[chunk].append((gap, m))
bygap[gap].append((chunk, m))
bycpu[cpu].append((gap, m))
from matplotlib import pyplot as plt
import seaborn as sns
sns.set_palette('Set1', len(groups))
for chunk, vals in sorted(bychunk.items()):
vals.sort()
xs, ys = zip(*vals)
plt.plot(xs, ys, label="chunk-size: %d" % chunk)
print chunk, vals
"""
for gap, vals in sorted(bygap.items()):
vals.sort()
xs, ys = zip(*vals)
plt.plot(xs, ys, label="gap-size: %d" % gap)
for cpu, vals in sorted(bycpu.items()):
vals.sort()
xs, ys = zip(*vals)
plt.plot(xs, ys, label="cpus: %d" % cpu)
"""
plt.xlabel("gap size")
#plt.xlabel("chunk size")
plt.ylabel("time (seconds)")
#plt.yscale('log', basey=2)
#plt.xscale('log', basex=10)
plt.legend()
plt.show()
1/0
for g, c in sorted(groups):
if g == '100' or c == '100': continue
print g, c, groups[(g, c)]
|
