Question

我的 CSV 文件：

YEAR,UTILITY_ID,UTILITY_NAME,OWNERSHIP,STATE_CODE,AMR_METERING_RESIDENTIAL,AMR_METERING_COMMERCIAL,AMR_METERING_INDUSTRIAL,AMR_METERING_TRANS,AMR_METERING_TOTAL,AMI_METERING_RESIDENTIAL,AMI_METERING_COMMERCIAL,AMI_METERING_INDUSTRIAL,AMI_METERING_TRANS,AMI_METERING_TOTAL,ENERGY_SERVED_RESIDENTIAL,ENERGY_SERVED_COMMERCIAL,ENERGY_SERVED_INDUSTRIAL,ENERGY_SERVED_TRANS,ENERGY_SERVED_TOTAL
2011,34,City of Abbeville - (SC),M,SC,880,14,,,894,,,,,,,,,,
2011,84,A & N Electric Coop,C,MD,135,25,,,160,,,,,,,,,,
2011,84,A & N Electric Coop,C,VA,31893,2107,0,,34000,,,,,,,,,,
2011,97,Adams Electric Coop,C,IL,8334,190,,,8524,,,,,0,,,,,0
2011,108,Adams-Columbia Electric Coop,C,WI,33524,1788,709,,36021,,,,,,,,,,
2011,118,Adams Rural Electric Coop, Inc,C,OH,7457,20,,,7477,,,,,,,,,,
2011,122,Village of Arcade,M,NY,3560,498,100,,4158,,,,,,,,,,
2011,155,Agralite Electric Coop,C,MN,4383,227,315,,4925,,,,,,,,,,

在此处下载 Spark 代码以读取 CSV 文件：

public class ReadFile8 {

    public static void main(String[] args) throws IOException {

        SparkSession session = new SparkSession.Builder().appName("CsvReader").master("local").getOrCreate();

        //Data taken by Local System
        Dataset<Row> file8Data = session.read().format("com.databricks.spark.csv").option("header", "true").load("file:///home/kumar/Desktop/Eletricaldata/file8_2011.csv");

        // Register the DataFrame as a SQL temporary view
        file8Data.createOrReplaceTempView("EletricalFile8Data");
        file8Data.show();
    }

}

如何使用 Java 在 Spark 中应用地图功能和平面地图功能？

Answer 1

您可以使用以下代码作为示例：

Dataset<Integer> years = file8Data.map((MapFunction<Row, Integer>) row -> row.<Integer>getAs("YEAR"), Encoders.INT());
Dataset<Integer> newYears = years.flatMap((FlatMapFunction<Integer, Integer>) year -> {
  return Arrays.asList(year + 1, year + 2).iterator();
}, Encoders.INT());

Answer 2

如果Encoders.INT（）无法正常工作，请尝试

Encoders $ .MODULE $ .INT（）

如何在spark java中的数据集上应用map函数

2 个答案: