pyspark报错，'DataFrame' object has no attribute '_jdf'

Question

dataZscore：

dataZscore = (dataTransformed - dataTransformed.mean(axis=0)) /dataTransformed.std(axis=0)
dataZscore.columns = ['Z' + i for i in dataTransformed.columns]
dataZscore.head()

    ZL     ZR     ZF     ZM     ZC
0     1.435707     -0.944948     14.034016     26.761154     0.315041
1     1.307152     -0.911894     9.073213     13.126864     0.315041
2     1.328381     -0.889859     8.718869     12.653481     0.315041
3     0.658476     -0.416098     0.781585     12.540622     0.315041
4     0.386032     -0.922912     9.923636     13.898736     0.315041

#Kmeans
#导入工具包
import time
import pyspark.ml.clustering as clu
# 获取程序运行前时间
start = time.time()
#通过上图观察可知最好的簇数量为5
#初始化Kmeans模型
kmodel = clu.KMeans(k=5, initMode='k-means||', initSteps=10,maxIter=300, seed=0)
kmodel.fit(dataZscore)
# # 获取程序运行结束后时间
# end = time.time()
# print(f"使用K-means聚类算法的运行时间为：%.3fs" % (end-start))
# # 聚类算法的评价指标CH值的计算：
# score = calinski_harabaz_score(dataZscore, labels)  # 至越大表示聚类效果越好
# print("CH值：", score)

一直报错，没办法解决

'DataFrame' object has no attribute '_jdf'

Answer 1

可以参考一下