使用scala从索引中获取TF-IDF值

单词和索引没有按顺序排列。例如文档0，

to

与

388

相关，但我不知道

333

与哪个单词相关。如何使用rawFeatures索引获取word。对于

CountVectorizer

，我可以使用

countVectorizerModel.词汇表

 import spark.implicits._
    val df = spark.sparkContext.parallelize(Array(
      (0, "to to Scala for better integration with Spark, and easier collaboration other".split(" ")),
      (1, "For example in the case when the document is mostly about".split(" ")),
      (2, "you need to to put some import declarations and create some data".split(" "))
    )).toDF("id", "content")

    val hashingTF = new HashingTF().setInputCol("content").setOutputCol("rawFeatures").setNumFeatures(2000)
    val featurizedData = hashingTF.transform(df)
    val idf = new IDF().setInputCol("rawFeatures").setOutputCol("features")
    val idfModel = idf.fit(featurizedData)
    val rescaleData = idfModel.transform(featurizedData)
    rescaleData.show(false)
+---+------------------------------------------------------------------------------------------+-------------------------------------------------------------------------------------------------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
|id |content                                                                                   |rawFeatures                                                                                            |features                                                                                                                                                                                                                                                                     |
+---+------------------------------------------------------------------------------------------+-------------------------------------------------------------------------------------------------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
|0  |[to, to, Scala, for, better, integration, with, Spark,, and, easier, collaboration, other]|(2000,[333,388,460,674,935,941,1036,1474,1534,1650,1988],[1.0,2.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0])|(2000,[333,388,460,674,935,941,1036,1474,1534,1650,1988],[0.28768207245178085,0.5753641449035617,0.6931471805599453,0.6931471805599453,0.6931471805599453,0.6931471805599453,0.6931471805599453,0.6931471805599453,0.6931471805599453,0.6931471805599453,0.6931471805599453])|
|1  |[For, example, in, the, case, when, the, document, is, mostly, about]                     |(2000,[342,956,1076,1243,1281,1445,1710,1760,1777,1820],[1.0,1.0,1.0,1.0,1.0,1.0,2.0,1.0,1.0,1.0])     |(2000,[342,956,1076,1243,1281,1445,1710,1760,1777,1820],[0.6931471805599453,0.6931471805599453,0.6931471805599453,0.6931471805599453,0.6931471805599453,0.6931471805599453,1.3862943611198906,0.6931471805599453,0.6931471805599453,0.6931471805599453])                     |
|2  |[you, need, to, to, put, some, import, declarations, and, create, some, data]             |(2000,[265,333,345,388,401,418,537,1400,1425,1695],[1.0,1.0,1.0,2.0,1.0,1.0,1.0,2.0,1.0,1.0])          |(2000,[265,333,345,388,401,418,537,1400,1425,1695],[0.6931471805599453,0.28768207245178085,0.6931471805599453,0.5753641449035617,0.6931471805599453,0.6931471805599453,0.6931471805599453,1.3862943611198906,0.6931471805599453,0.6931471805599453])                         |
+---+------------------------------------------------------------------------------------------+-------------------------------------------------------------------------------------------------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+