Problem with feature importance decision tree

Feature importance Decision Tree

# Plot importance of variables
feature_importance = model.feature_importances_ 
sorted_idx = np.argsort(feature_importance) # Sort index on feature importance
fig = plt.figure(figsize=(20, 15)) # Set plot size (denoted in inches)
plt.barh(range(len(sorted_idx)), feature_importance[sorted_idx], align='center')
plt.yticks(range(len(sorted_idx)), np.array(X_test.columns)[sorted_idx])

plt.xlabel("Feature importance") # Add x axis
plt.ylabel("Feature") # Add y axis