BigBench/plot_results.py at main · BreezeWhite/BigBench · GitHub

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
#!/usr/bin/env python
import json
import matplotlib.pyplot as plt
import numpy as np

# Load hyperfine results
with open("results.json", encoding="utf-8") as f:
    data = json.load(f)

results = data["results"]

command_to_label = {
    "raw-bbp": "raw f64",
    "rs-decimal-bbp": "rust-decimal",
    "big-decimal-bbp": "bigdecimal",
    "rug-bbp": "rug",
    "dashu-bbp": "dashu",
    "big-float-bbp": "num-bigfloat",
    "astro-float-bbp": "astro-float",
    "fastnum-bbp": "fastnum",
    "decimal-rs-bbp": "decimal-rs",
    "prim-fpdec-bbp": "prim-fpdec",
    "malachite-bbp": "malachite",
    "fixed-num-bbp": "fixed-num",
    "python main.py C 1000": "Python C",
    "python main.py PY 1000": "Python PY"
}

correctness_map = {
    "raw f64": 16,
    "rust-decimal": 28,
    "fixed-num": 19,
    "bigdecimal": 1000,
    "rug": 1000,
    "dashu": 1000,
    "num-bigfloat": 39,
    "astro-float": 1000,
    "fastnum": 307,
    "decimal-rs": 37,
    "prim-fpdec": 35,
    "malachite": 1000,
    "Python C": 1000,
    "Python PY": 1000
}

plot_data = []
for r in results:
    cmd = r["command"]
    label = cmd
    for k, v in command_to_label.items():
        if k in cmd:
            label = v
            break
    median_ms = r["median"] * 1000.0
    corr = correctness_map.get(label, 0)
    plot_data.append({
        "label": label,
        "median_ms": median_ms,
        "correctness": corr
    })

# Sort by median time (fastest to slowest)
plot_data.sort(key=lambda x: x["median_ms"])

labels = [x["label"] for x in plot_data]
medians = [x["median_ms"] for x in plot_data]
correctness_vals = [x["correctness"] for x in plot_data]

# Styling
plt.style.use('seaborn-v0_8-whitegrid' if 'seaborn-v0_8-whitegrid' in plt.style.available else 'default')

fig, ax1 = plt.subplots(figsize=(16, 10), dpi=150)

x = np.arange(len(labels))  # the label locations
width = 0.35  # the width of the bars

# Colors
color_time = '#2b5c8f'
color_correct = '#d95f02'

# Primary axis: Time (ms)
rects1 = ax1.bar(x - width/2, medians, width, label='Runtime (ms)', color=color_time, edgecolor='none')
ax1.set_ylabel('Median Runtime (ms)', color=color_time, fontsize=16, fontweight='bold')
ax1.tick_params(axis='y', labelcolor=color_time, labelsize=12)

# Secondary axis: Correctness (digits)
ax2 = ax1.twinx()
ax2.grid(False)  # Turn off grid for the secondary y-axis to prevent clutter
rects2 = ax2.bar(x + width/2, correctness_vals, width, label='Correctness (digits)', color=color_correct, edgecolor='none')
ax2.set_ylabel('Correctness (digits)', color=color_correct, fontsize=16, fontweight='bold')
ax2.tick_params(axis='y', labelcolor=color_correct, labelsize=12)

ax1.set_title("BBP Pi Calculation Benchmark", fontsize=22, fontweight='bold', pad=25)
ax1.set_xticks(x)
ax1.set_xticklabels(labels, rotation=45, ha='right', fontsize=13, fontweight='bold')

# Add values on top of the bars with size 12 font, rotated 90 degrees
max_median = max(medians)
padding_y1 = max_median * 0.02
for rect in rects1:
    height = rect.get_height()
    if height >= 1000:
        label_text = f"{height/1000:.2f} s"
    else:
        label_text = f"{height:.1f} ms"
    ax1.text(
        rect.get_x() + rect.get_width()/2.0,
        height + padding_y1,
        label_text,
        ha='center',
        va='bottom',
        fontsize=12,
        fontweight='bold',
        color=color_time,
        rotation=90
    )

max_correct = max(correctness_vals)
padding_y2 = max_correct * 0.02
for rect in rects2:
    height = rect.get_height()
    label_text = f"{int(height)}"
    ax2.text(
        rect.get_x() + rect.get_width()/2.0,
        height + padding_y2,
        label_text,
        ha='center',
        va='bottom',
        fontsize=12,
        fontweight='bold',
        color=color_correct,
        rotation=90
    )

# Legend
lines1, labels1 = ax1.get_legend_handles_labels()
lines2, labels2 = ax2.get_legend_handles_labels()
ax1.legend(lines1 + lines2, labels1 + labels2, loc='upper left', fontsize=12)

# Set ylimits with extra space for vertical labels
ax1.set_ylim(ymin=0.0, ymax=max_median * 1.35)
ax2.set_ylim(ymin=0.0, ymax=max_correct * 1.35)

plt.tight_layout()
plt.savefig("results.png", bbox_inches='tight', dpi=300)
print("Plot saved successfully to results.png")