Database Query Performance ANOVA

import matplotlib.pyplot as plt
import numpy as np
import scipy.stats as stats

np.random.seed(123)

# Query execution time in milliseconds
postgresql = np.random.lognormal(2.5, 0.4, 150)
mysql = np.random.lognormal(2.7, 0.45, 150)
mongodb = np.random.lognormal(2.3, 0.35, 150)
redis = np.random.lognormal(1.2, 0.3, 150)

# Clip for visualization
postgresql = np.clip(postgresql, 2, 80)
mysql = np.clip(mysql, 2, 100)
mongodb = np.clip(mongodb, 2, 60)
redis = np.clip(redis, 0.5, 15)

F_stat, p_value = stats.f_oneway(postgresql, mysql, mongodb, redis)

fig, ax = plt.subplots(figsize=(12, 7), facecolor='#0a0a0f')
ax.set_facecolor('#0a0a0f')

colors = ['#276CF5', '#F5B027', '#6CF527', '#F5276C']
data = [postgresql, mysql, mongodb, redis]

bp = ax.boxplot(data, positions=[1, 2, 3, 4], widths=0.6, patch_artist=True,
                medianprops={'color': 'white', 'linewidth': 2},
                whiskerprops={'color': '#555555', 'linewidth': 1.5},
                capprops={'color': '#555555', 'linewidth': 1.5},
                flierprops={'marker': '.', 'markerfacecolor': '#666666', 'markersize': 5})

for patch, color in zip(bp['boxes'], colors):
    patch.set_facecolor(color)
    patch.set_alpha(0.7)
    patch.set_edgecolor('white')
    patch.set_linewidth(1.5)

labels = ['PostgreSQL', 'MySQL', 'MongoDB', 'Redis']

# Percentile annotations
for i, (d, color) in enumerate(zip(data, colors)):
    p50 = np.percentile(d, 50)
    p99 = np.percentile(d, 99)
    ax.text(i+1, -8, f'P50:{p50:.1f} P99:{p99:.1f}', ha='center', fontsize=8, color=color)

# Performance threshold
ax.axhline(y=50, color='#ef4444', linestyle='--', alpha=0.6, linewidth=1.5)
ax.text(4.45, 50, 'SLA', fontsize=8, color='#ef4444', va='center')

# Stats header
stats_text = f"ANOVA: F={F_stat:.1f}, p<0.001 | Fastest: Redis (μ={redis.mean():.1f}ms)"
bbox = dict(boxstyle="round,pad=0.3", facecolor='#020B14', edgecolor='#F5276C', lw=2)
ax.text(0.5, 1.02, stats_text, transform=ax.transAxes, fontsize=10, color='white',
        ha='center', va='bottom', fontfamily='monospace', bbox=bbox)

ax.set_xticks([1, 2, 3, 4])
ax.set_xticklabels(labels, fontsize=11, color='white')
ax.set_ylabel('Query Time (ms)', fontsize=12, color='white', fontweight='500')
ax.set_title('Database Engine Performance Comparison\nSELECT Query Benchmark', 
             fontsize=14, color='white', fontweight='bold', pad=25)

ax.tick_params(colors='#888888')
for spine in ax.spines.values():
    spine.set_color('#333333')
ax.yaxis.grid(True, color='#1a1a2e', linewidth=0.5)
ax.set_axisbelow(True)
ax.set_ylim(-15, 110)

plt.tight_layout()
plt.show()