51 lines
1.6 KiB
Python
51 lines
1.6 KiB
Python
|
import pandas as pd
|
||
|
|
import matplotlib.pyplot as plt
|
||
|
|
import numpy as np
|
||
|
|
|
||
|
|
df = pd.read_csv('res.csv')
|
||
|
|
|
||
|
|
grouped = df.groupby(['Structure', 'Mode']).agg({
|
||
|
|
'Insert (sec)': ['mean', 'std'],
|
||
|
|
'Search (sec)': ['mean', 'std'],
|
||
|
|
'Delete (sec)': ['mean', 'std']
|
||
|
|
}).reset_index()
|
||
|
|
|
||
|
|
grouped.columns = ['Structure', 'Mode', 'Insert_mean', 'Insert_std',
|
||
|
|
'Search_mean', 'Search_std', 'Delete_mean', 'Delete_std']
|
||
|
|
|
||
|
|
structures = grouped['Structure'].unique()
|
||
|
|
modes = grouped['Mode'].unique()
|
||
|
|
|
||
|
|
fig, axes = plt.subplots(1, 3, figsize=(15, 5))
|
||
|
|
operations = ['Insert', 'Search', 'Delete']
|
||
|
|
colors = {'random': 'skyblue', 'sorted': 'lightcoral'}
|
||
|
|
|
||
|
|
for i, op in enumerate(operations):
|
||
|
|
ax = axes[i]
|
||
|
|
mean_col = f'{op}_mean'
|
||
|
|
std_col = f'{op}_std'
|
||
|
|
x = np.arange(len(structures))
|
||
|
|
width = 0.35
|
||
|
|
for j, mode in enumerate(modes):
|
||
|
|
data = grouped[grouped['Mode'] == mode]
|
||
|
|
means = [data[data['Structure'] == s][mean_col].values[0] for s in structures]
|
||
|
|
stds = [data[data['Structure'] == s][std_col].values[0] for s in structures]
|
||
|
|
offset = (j - 0.5) * width
|
||
|
|
bars = ax.bar(x + offset, means, width, yerr=stds, capsize=3,
|
||
|
|
label=mode.capitalize(), color=colors[mode])
|
||
|
|
|
||
|
|
ax.set_xticks(x)
|
||
|
|
ax.set_xticklabels(structures)
|
||
|
|
ax.set_ylabel('Time (seconds)')
|
||
|
|
ax.set_title(f'{op} Time')
|
||
|
|
ax.legend()
|
||
|
|
|
||
|
|
if op == 'Insert':
|
||
|
|
ax.set_yscale('log')
|
||
|
|
ax.set_ylabel('Time (seconds) [log scale]')
|
||
|
|
|
||
|
|
plt.tight_layout()
|
||
|
|
plt.savefig('performance_plots.png', dpi=150)
|
||
|
|
plt.show()
|
||
|
|
|
||
|
|
print("Графики сохранены в файл performance_plots.png")
|