大数据IMF传奇行动绝密课程第58课：使用Java和Scala在IDE中开发DataFrame实战

使用Java和Scala在IDE中开发DataFrame实战

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