Scala Spark中的RDD过滤器

Question

Scala Spark中的RDD过滤器

8

我有一个数据集，想要提取那些 (review/time) 在 x 和 y 之间的 (review/text)，例如(1183334400 < time < 1185926400)，

以下是我的部分数据：

product/productId: B000278ADA
product/title: Jobst Ultrasheer 15-20 Knee-High Silky Beige Large
product/price: 46.34
review/userId: A17KXW1PCUAIIN
review/profileName: Mark Anthony "Mark"
review/helpfulness: 4/4
review/score: 5.0
review/time: 1174435200
review/summary: Jobst UltraSheer Knee High Stockings
review/text: Does a very good job of relieving fatigue.

product/productId: B000278ADB
product/title: Jobst Ultrasheer 15-20 Knee-High Silky Beige Large
product/price: 46.34
review/userId: A9Q3932GX4FX8
review/profileName: Trina Wehle
review/helpfulness: 1/1
review/score: 3.0
review/time: 1352505600
review/summary: Delivery was very long wait.....
review/text: It took almost 3 weeks to recieve the two pairs of stockings .

product/productId: B000278ADB
product/title: Jobst Ultrasheer 15-20 Knee-High Silky Beige Large
product/price: 46.34
review/userId: AUIZ1GNBTG5OB
review/profileName: dgodoy
review/helpfulness: 1/1
review/score: 2.0
review/time: 1287014400
review/summary: sizes recomended in the size chart are not real
review/text: sizes are much smaller than what is recomended in the chart. I tried to put it and sheer it!.

我的Spark-Scala代码：

import org.apache.hadoop.conf.Configuration
import org.apache.hadoop.io.{LongWritable, Text}
import org.apache.hadoop.mapreduce.lib.input.TextInputFormat
import org.apache.spark.{SparkConf, SparkContext}

object test1 {
  def main(args: Array[String]): Unit = {
    val conf1 = new SparkConf().setAppName("golabi1").setMaster("local")
    val sc = new SparkContext(conf1)
    val conf: Configuration = new Configuration
    conf.set("textinputformat.record.delimiter", "product/title:")
    val input1=sc.newAPIHadoopFile("data/Electronics.txt",     classOf[TextInputFormat], classOf[LongWritable], classOf[Text], conf)
    val lines = input1.map { text => text._2}
    val filt = lines.filter(text=>(text.toString.contains(tt => tt in (startdate until enddate))))
    filt.saveAsTextFile("data/filter1")
  }
}

但是我的代码运行不好，

我该如何筛选这些行？

- Esmaeil zahedi

1

我在您的输入文件中没有看到分隔符字符串“product / productId：”。 - Nikita

1

你期望什么输出结果？你遇到了什么问题？ - maasg

1个回答

网页内容由stack overflow 提供, 点击上面的

可以查看英文原文，
原文链接

- Daniel Langdon · Accepted Answer

这其实很简单。试试这个方法：

object test1 
{
  def main(args: Array[String]): Unit = 
  {
    val conf1 = new SparkConf().setAppName("golabi1").setMaster("local")
    val sc = new SparkContext(conf1)

    def extractDateAndCompare(line: String): Boolean=
    {
        val from = line.indexOf("/time: ") + 7
        val to = line.indexOf("review/text: ") -1
        val date = line.substring(from, to).toLong
        date > startDate && date < endDate
    }

    sc.textFile("data/Electronics.txt")
        .filter(extractDateAndCompare)
        .saveAsTextFile("data/filter1")
  }
}

我通常发现那些中间辅助方法可以使事情更清晰。当然，这假设边界日期在某处被定义并且输入文件包含格式问题。我故意这样做是为了保持简单, 但是如果你遇到错误，添加try、返回Option子句和使用flatMap()可以帮助你避免错误。

另外，您的原始文本有点繁琐，您可能需要探索Json、TSV文件或其他一些更简单的格式。