Model Feature Importance

import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd
import numpy as np
from matplotlib.colors import LinearSegmentedColormap

np.random.seed(42)

# Features and models
features = ['Age', 'Income', 'Education', 'Location', 'Tenure', 
            'Usage', 'Support_Calls', 'Payment_Delay', 'Contract', 'Satisfaction']
models = ['Random Forest', 'XGBoost', 'LightGBM', 'CatBoost', 'Neural Net']

# Generate importance scores
importance = np.random.rand(len(features), len(models))
# Add structure - some features more important across models
importance[0, :] += 0.3  # Age important
importance[5, :] += 0.25  # Usage important
importance[7, :] += 0.2  # Payment_Delay important
importance = importance / importance.sum(axis=0)  # Normalize

df = pd.DataFrame(importance, index=features, columns=models)

# NEON cyan-magenta colormap
neon_cm = LinearSegmentedColormap.from_list('neon_cm', ['#ffffff', '#27D3F5', '#4927F5', '#F5276C'])

fig, ax = plt.subplots(figsize=(9, 6), facecolor='#ffffff')
ax.set_facecolor('#ffffff')

sns.heatmap(df, cmap=neon_cm, annot=True, fmt='.1%', linewidths=2, linecolor='#ffffff',
            annot_kws={'size': 9, 'fontweight': '500'},
            cbar_kws={'shrink': 0.8, 'label': 'Importance'}, ax=ax)

ax.set_title('Feature Importance Comparison', color='#1f2937', fontsize=14, fontweight='bold', pad=15)
ax.set_xlabel('Model', color='#1f2937', fontsize=11)
ax.set_ylabel('Feature', color='#1f2937', fontsize=11)
ax.tick_params(colors='#374151', labelsize=10)
ax.set_xticklabels(ax.get_xticklabels(), rotation=30, ha='right')

plt.tight_layout()
plt.show()