Spark Pipeline取出中间步骤的模型参数

通过pipelinemodel的stages可以取出中间步骤的输出结果

val model = pipeline.fit(featureDF)
model.stages(7).asInstanceOf[KMeansModel].clusterCenters

val kmeans = new KMeans()
      .setInitMode("k-means||")
      .setFeaturesCol("pcaFeatures")
      .setK(K)
      .setMaxIter(3)
      .setSeed(1L)
      .setTol(0.01)
    
val pipeline = new Pipeline()
      .setStages(Array(method_indexer, status_indexer, agent_indexer,content_type_indexer,tld_indexer,assembler,pca,kmeans))

val model = pipeline.fit(featureDF)

//取出kmeans模型的参数
//model.stages(7).asInstanceOf[KMeansModel].clusterCenters