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
|
from __future__ import annotations
import matplotlib.pyplot as plt
import numpy as np
# Res 0.1 31776, 0.05 234546, 0.025 1801086, 0.02 3488856, 0.0175 5147961,0.015 7960200
dofs = [31776, 234546, 1801086, 3488856, 5147961, 7960200]
def visualize_side_by_side(dofs):
fig, ax = plt.subplots()
plt.grid("on", zorder=1)
procs = []
first = True
slaves = []
totals = np.zeros(len(dofs))
for i, dof in enumerate(dofs):
infile = open("results_bench_{0:d}.txt".format(dof), "r")
# Read problem info
procs.append(int(infile.readline().split(": ")[-1].strip("\n")))
# Skip num dofs
infile.readline()
slaves.append(int(infile.readline().split(": ")[-1].strip("\n")))
solve_iterations = int(infile.readline().split(": ")[-1].strip("\n"))
# Skip info line
infile.readline()
# Read timings
operations = infile.readlines()
colors = [
"tab:blue",
"tab:brown",
"tab:orange",
"tab:green",
"tab:red",
"tab:purple",
"tab:cyan",
"tab:olive",
]
total_time = 0
for j, line in enumerate(operations):
data = line.strip("\n").split(" ")
if data[0] == "Backsubstitution":
continue
if first:
plt.bar(
i,
float(data[2]),
0.5,
bottom=total_time,
label=data[0],
color=colors[j],
zorder=2,
)
else:
plt.bar(i, float(data[2]), 0.5, bottom=total_time, color=colors[j], zorder=2)
if data[0] == "Solve":
ax.annotate(
"{0:d}".format(solve_iterations),
xy=(i, total_time + float(data[2]) / 2),
xytext=(0, 0),
textcoords="offset points",
ha="center",
va="bottom",
)
total_time += float(data[2])
first = False
totals[i] = total_time
ax.set_xticks(range(len(dofs)))
labels = dofs
ax.set_xticklabels(labels)
assert np.allclose(procs, procs[0])
# Shrink current axis by 20%
box = ax.get_position()
ax.set_position([box.x0, box.y0, box.width * 0.8, box.height])
ax.legend(bbox_to_anchor=(0.95, 0.5))
plt.xlabel("Degrees of freedom")
plt.title(f"Average runtime of operations with {procs[0]} MPI ranks")
plt.ylabel("Runtime (s)")
plt.savefig("comparison_bars.png")
# Second figure
plt.figure()
plt.ylabel("Runtime (s)")
plt.xlabel("Degrees of freedom")
power_min = int(np.log10(min(dofs)))
power_max = int(np.log10(max(dofs)) + 1)
power_tmin = int(np.log10(min(totals)) - 1)
power_tmax = int(np.log10(max(totals)) + 1)
plt.axis((10**power_min, 10**power_max, 10**power_tmin, 10**power_tmax))
plt.plot(dofs, totals, "-ro", label="Simulations")
# 7
plt.plot([3.1 * 10**4, 3.1 * 10**7], [4.3 * 10**-1, 4.3 * 10**2], "--g", label="Order 1")
# 6 procs
# plt.plot([3.1 * 10**4, 3.1 * 10**7], [4.8 * 10**-1, 4.8 * 10**2], "--g", label="Order 1")
# 4 procs
# plt.plot([3.1 * 10**4, 3.1 * 10**7], [5.7 * 10**-1, 5.7 * 10**2], "--g", label="Order 1")
plt.title(f"Total runtime of core operations with {procs[0]} MPI ranks")
plt.legend()
plt.grid()
plt.xscale("log")
plt.yscale("log")
plt.savefig("comparison.png")
def visualize_single(dof):
infile = open("results_bench_{0:d}.txt".format(dof), "r")
# Read problem info
procs = infile.readline().split(": ")[-1].strip("\n")
# Skip num dofs
infile.readline()
slaves = int(infile.readline().split(": ")[-1].strip("\n"))
infile.readline()
# solve_iterations = int(infile.readline().split(": ")[-1].strip("\n"))
# Skip info line
infile.readline()
# Read timings
operations = infile.readlines()
colors = [
"tab:blue",
"tab:brown",
"tab:orange",
"tab:green",
"tab:red",
"tab:purple",
"tab:cyan",
"tab:olive",
]
fig, ax = plt.subplots()
plt.grid("on", zorder=1)
for j, line in enumerate(operations):
data = line.strip("\n").split(" ")
if data[0] == "Backsubstitution":
continue
plt.bar(j, float(data[2]), 0.5, label=data[0], color=colors[j], zorder=2)
ax.set_xticks([])
# ax.set_yscale("log")
plt.legend()
plt.ylabel("Runtime (s)")
# Shrink current axis by 20%
# box = ax.get_position()
# ax.set_position([box.x0, box.y0, box.width * 0.8, box.height])
# ax.legend(bbox_to_anchor=(0.95, 0.5))
plt.title(f"Average runtime of operations with {dof} ({slaves} slaves) on {procs} MPI ranks")
plt.savefig(f"comparison_{slaves}.png")
visualize_single(max(dofs))
visualize_side_by_side(dofs)
|
