1 package big.data.analyse.sparksql 2  3 import org.apache.spark.sql.{Row, SparkSession} 4 import org.apache.spark.sql.types.{IntegerType, StringType, StructField, StructType} 5  6 /** 7   * Created by zhen on 2018/11/8. 8   */ 9 object SparkInFuncation {10   def main(args: Array[String]) {11     val spark = SparkSession.builder().appName("spark内置函数")12       .master("local[2]").getOrCreate()13     val userData = Array(14       "2015,1,www.baidu.com",15       "2016,4,www.google.com",16       "2017,3,www.apache.com",17       "2015,6,www.spark.com",18       "2016,2,www.hadoop.com",19       "2017,8,www.solr.com",20       "2017,4,www.hive.com"21     )22     val sc = spark.sparkContext23     val sqlContext = spark.sqlContext24     val userDataRDD = sc.parallelize(userData) // 转化为RDD25     val userDataType = userDataRDD.map(line => {26       val Array(age, id, url) = line.split(",")27       Row(28         age, id.toInt, url29       )30     })31     val structTypes = StructType(Array(32       StructField("age", StringType, true),33       StructField("id", IntegerType, true),34       StructField("url", StringType, true)35     ))36     // RDD转化为DataFrame37     val userDataFrame = sqlContext.createDataFrame(userDataType,structTypes)38 39     import org.apache.spark.sql.functions._40     userDataFrame41       .groupBy("age") // 分组42       .agg(("id","sum")) // 求和43       .orderBy(desc("age")) // 排序44       .show()45   }46 }

答案：