绘制K-means散点图时报错'DataFrame' object has no attribute 'label'如何解决

这个报错是两个问题叠加导致的：

1. 执行完K-means聚类后，你没有将生成的聚类标签写入DataFrame的label列，调用df.label时pandas找不到对应属性
2. 现有DataFrame的消费列名是PURCHASES，绘图代码里写的PURCHASE少了末尾的S，后续也会触发列不存在的报错

• 第一步：将K-means的聚类结果写入DataFrame
  假设你已经完成K-means模型训练，执行以下代码新增label列：

  df['label'] = kmeans.labels_
  

• 第二步：修正绘图代码里的列名错误，把所有df["PURCHASE"]改成df["PURCHASES"]
• 第三步：重新运行绘图代码即可

plt.scatter(df["BALANCE"][df.label == 0], df["PURCHASES"][df.label == 0],s=80,c='magenta',label='Careful')
plt.scatter(df["BALANCE"][df.label == 1], df["PURCHASES"][df.label == 1],s=80,c='yellow',label='Standard')
plt.scatter(df["BALANCE"][df.label == 2], df["PURCHASES"][df.label == 2],s=80,c='green',label='Target')
plt.scatter(df["BALANCE"][df.label == 3], df["PURCHASES"][df.label == 3],s=80,c='cyan',label='Careless')
plt.scatter(df["BALANCE"][df.label == 4], df["PURCHASES"][df.label == 4],s=80,c='burlywood',label='Sensible')
plt.scatter(kmeans.cluster_centers_[:, 0], kmeans.cluster_centers_[:, 1], s=300, c='red', label = 'Centroids')
plt.title('Customers Segmentation based on their Credit Card usage bhaviour')
plt.xlabel('BALANCE')
plt.ylabel('PURCHASES')
plt.legend()
plt.show()

内容的提问来源于stack exchange，提问作者Mohamed Guidoum