1、使用Java开发DataFrame
2、使用Scala开发DataFrame
创建DataFrame的时候可以来自于其它RDD,来源于Hive表,以及其他数据来源,例如json文件
SQLContext只支持SQL一种方言(delax?),HiveContext支持SQL方言以及其它方言,通过设置都可以支持。
//F:\sparkData\people.json文件{"name":"Michael"}{"name":"Andy","age":31}{"name":"Justin","age":20}
一、使用Java开发DataFrame
package com.tom.spark.SparkApps.sql;import org.apache.spark.SparkConf;import org.apache.spark.api.java.JavaSparkContext;import org.apache.spark.sql.DataFrame;import org.apache.spark.sql.SQLContext;/** * */public class DataFrameOps { /** * @param args */ public static void main(String[] args) { //创建SparkConf用于读取系统配置信息并设置当前应用程序的名字 SparkConf conf = new SparkConf().setAppName("DataFrameOps").setMaster("local"); //创建JavaSparkContext对象实例作为整个Driver的核心基石 JavaSparkContext sc = new JavaSparkContext(conf); //设置日志级别为WARN sc.setLogLevel("WARN"); //创建SQLContext上下文对象用于SQL的分析 SQLContext sqlContext = new SQLContext(sc); //创建Data Frame,可以简单的认为DataFrame是一张表 DataFrame df = sqlContext.read().json("F:\\sparkData\\people.json"); //select * from table df.show(); //desc table df.printSchema(); //select name from table df.select(df.col("name")).show(); //select name, age+10 from table df.select(df.col("name"), df.col("age").plus(10)).show(); //select * from table where age > 21 df.filter(df.col("age").gt(21)).show(); //select age, count(1) from table group by age df.groupBy("age").count().show(); //df.groupBy(df.col("age")).count().show(); }}
以下为程序输出:
+----+-------+| age| name|+----+-------+|null|Michael|| 31| Andy|| 20| Justin|+----+-------+root |-- age: long (nullable = true) |-- name: string (nullable = true)+-------+| name|+-------+|Michael|| Andy|| Justin|+-------++-------+----------+| name|(age + 10)|+-------+----------+|Michael| null|| Andy| 41|| Justin| 30|+-------+----------++---+----+|age|name|+---+----+| 31|Andy|+---+----++----+-----+| age|count|+----+-----+| 31| 1||null| 1|| 20| 1|+----+-----+
二、使用Scala开发DataFrame
package com.tom.spark.sqlimport org.apache.spark.sql.SQLContextimport org.apache.spark.{SparkConf, SparkContext}/** * */object DataFrameOps { def main(args: Array[String]): Unit = { val conf = new SparkConf().setAppName("DataFrameOps").setMaster("local") val sc = new SparkContext(conf) sc.setLogLevel("WARN") val sqlContext = new SQLContext(sc) val df = sqlContext.read.json("F:\\sparkData\\people.json") df.show() df.printSchema() df.select("name").show() df.select(df("name"),df("age")+10).show() df.filter(df("age")>21).show() df.groupBy("age").count().show() }}
以下为程序输出
+----+-------+| age| name|+----+-------+|null|Michael|| 31| Andy|| 20| Justin|+----+-------+root |-- age: long (nullable = true) |-- name: string (nullable = true)+-------+| name|+-------+|Michael|| Andy|| Justin|+-------++-------+----------+| name|(age + 10)|+-------+----------+|Michael| null|| Andy| 41|| Justin| 30|+-------+----------++---+----+|age|name|+---+----+| 31|Andy|+---+----++----+-----+| age|count|+----+-----+| 31| 1||null| 1|| 20| 1|+----+-----+
spark-submit可以指定–file参数,可以把hive-site.xml中指定的hive文件夹添加进来
spark-submit --class com.dt.spark.sql.DataFrameOps --files /usr/local/hive/apache-hive-1.2.1-bin/conf/hive-site.xml --driver-class-path /usr/local/hive/apace-hive-1.2.1-bin/mysql-connector-java-5.1.35-bin.jar --master spark://Master:7077 /root/Documents/SparkApps/WordCount.jar
联系客服