Feature Importance Heatmap

import matplotlib.pyplot as plt
import numpy as np
from matplotlib.colors import LinearSegmentedColormap
from matplotlib.patches import FancyBboxPatch

fig, ax = plt.subplots(figsize=(12, 9), facecolor='#ffffff')
ax.set_facecolor('#ffffff')

np.random.seed(42)
features = ['Age', 'Income', 'Education', 'Location', 'Device', 'Time', 'History', 'Referral']
models = ['XGBoost', 'Random Forest', 'Neural Net', 'LogReg', 'SVM']
data = np.random.uniform(0, 1, (len(models), len(features)))

colors = ['#1e293b', '#164e63', '#0891b2', '#f472b6', '#fce7f3']
cmap = LinearSegmentedColormap.from_list('ml', colors, N=256)

cell_w, cell_h = 0.88, 0.82
for i in range(len(models)):
    for j in range(len(features)):
        val = data[i, j]
        rect = FancyBboxPatch((j - cell_w/2, i - cell_h/2), cell_w, cell_h,
                               boxstyle="round,pad=0.02,rounding_size=0.12",
                               facecolor=cmap(val), edgecolor='#e2e8f0', linewidth=1.5)
        ax.add_patch(rect)
        ax.text(j, i, f'{val:.2f}', ha='center', va='center', color='#1e293b', fontsize=9, fontweight='bold')

ax.set_xlim(-0.5, len(features)-0.5)
ax.set_ylim(-0.5, len(models)-0.5)
ax.set_aspect('equal')
ax.invert_yaxis()
ax.set_xticks(range(len(features)))
ax.set_yticks(range(len(models)))
ax.set_xticklabels(features, rotation=45, ha='right', color='#64748b', fontsize=10, fontweight='500')
ax.set_yticklabels(models, color='#1e293b', fontsize=10, fontweight='500')

sm = plt.cm.ScalarMappable(cmap=cmap, norm=plt.Normalize(0, 1))
cbar = plt.colorbar(sm, ax=ax, shrink=0.8, pad=0.02)
cbar.set_label('Importance', color='#1e293b', fontsize=11)
cbar.outline.set_edgecolor('#e2e8f0')
plt.setp(plt.getp(cbar.ax.axes, 'yticklabels'), color='#64748b')

for spine in ax.spines.values(): spine.set_visible(False)
ax.tick_params(length=0)
ax.set_title('Feature Importance by Model', fontsize=18, color='#1e293b', fontweight='bold', pad=20)
plt.tight_layout()
plt.show()