搭好的Spark当然要先写一个最简单的WordCount练练手。 那么,需求是: 1、统计Spark下README.md文件的词频; 2、输出较多,筛选出现次数超过10次的,词频逆序显示
注意:
[]
,括号中的数字表示执行任务的线程数;
4.3 local[*] 表示CPU有几个核就用几个线程。这里可以直接构建scala工程,但是一次准备写scala和java的,因此从基础的maven项目开始构建。
略
image.png
image.png
构建scala工程
package com.junzerg
import org.apache.spark.rdd.RDD
import org.apache.spark.{SparkConf, SparkContext}
object WordCount {
def main(args: Array[String]): Unit = {
/**
* 获取编程入口
*/
val conf: SparkConf = new SparkConf()
conf.setAppName("WordCount")
conf.setMaster("local")
val sc: SparkContext = new SparkContext(conf)
/**
* 通过编程入口加载数据
*/
val linesRdd:RDD[String] = sc.textFile("/Users/pengjunzhe/Downloads/spark-2.4.0-bin-hadoop2.7/README.md")
/**
* 对数据进行处理
*/
val wordCountRdd = linesRdd
.flatMap(_.split(" "))
.map((_, 1))
.reduceByKey(_ + _)
.filter(_._2 > 10)
.filter(_._1 != "")
.sortBy(_._2, false)
/**
* 对结果数据进行处理l
*/
wordCountRdd.foreach(println(_))
sc.stop()
}
}
首次运行报错,百度又是一通鬼扯:
Exception in thread "main" java.lang.ArrayIndexOutOfBoundsException: 10582
打断点,看异常栈,发现最后抛出的异常在:
declaringClass = "com.thoughtworks.paranamer.BytecodeReadingParanamer$ClassReader"
methodName = "accept"
fileName = "BytecodeReadingParanamer.java"
lineNumber = 563
这个类默认下载了2.7版本。和jdk8兼容不是很好,要去pom指定一下:
<dependency>
<groupId>com.thoughtworks.paranamer</groupId>
<artifactId>paranamer</artifactId>
<version>2.8</version>
</dependency>
在IDEA和pom中,退一下版本
/**
* @author pengjunzhe
*/
public class WordCountJdk7 {
private static final Pattern SPACE = Pattern.compile(" ");
public static void main(String[] args) {
/**
* 获取编程入口
*/
SparkConf conf = new SparkConf();
conf.setMaster("local");
conf.setAppName("WordCountJdk7");
JavaSparkContext sc = new JavaSparkContext(conf);
/**
* 通过编程入口加载数据
*/
JavaRDD<String> lines = sc.textFile("/Users/pengjunzhe/Downloads/spark-2.4.0-bin-hadoop2.7/README.md", 2);
/**
* 对数据进行处理
*/
JavaRDD<String> words = lines.flatMap(new FlatMapFunction<String, String>() {
@Override
public Iterator<String> call(String s) {
return Arrays.asList(s.split(" ")).iterator();
}
});
JavaPairRDD<String, Integer> ones = words.mapToPair(new PairFunction<String, String, Integer>() {
@Override
public Tuple2<String, Integer> call(String s) {
return new Tuple2<>(s, 1);
}
});
ones = ones.reduceByKey(new Function2<Integer, Integer, Integer>() {
@Override
public Integer call(Integer v1, Integer v2) throws Exception {
return v1 + v2;
}
}).filter(new Function<Tuple2<String, Integer>, Boolean>() {
@Override
public Boolean call(Tuple2<String, Integer> v1) throws Exception {
return v1._2 > 10;
}
}).filter(new Function<Tuple2<String, Integer>, Boolean>() {
@Override
public Boolean call(Tuple2<String, Integer> v1) throws Exception {
return v1._1.length() > 1;
}
});
JavaPairRDD<Integer, String> transdOnes = ones.mapToPair(new PairFunction<Tuple2<String, Integer>, Integer, String>() {
@Override
public Tuple2<Integer, String> call(Tuple2<String, Integer> t2) throws Exception {
return new Tuple2<>(t2._2, t2._1);
}
}).sortByKey();
ones = transdOnes.mapToPair(new PairFunction<Tuple2<Integer, String>, String, Integer>() {
@Override
public Tuple2<String, Integer> call(Tuple2<Integer, String> t2) throws Exception {
return new Tuple2<>(t2._2, t2._1);
}
});
/**
* 对结果数据进行处理
*/
ones.foreach(new VoidFunction<Tuple2<String, Integer>>() {
@Override
public void call(Tuple2<String, Integer> t2) throws Exception {
System.out.println(t2);
}
});
/**
* 结束关闭入口
*/
sc.stop();
}
}
/**
* @author pengjunzhe
*/
public class WordCountJdk8 {
public static void main(String[] args) {
/**
* 获取编程入口
*/
SparkConf conf = new SparkConf();
conf.setMaster("local").setAppName("WordCountJdk8");
JavaSparkContext sc = new JavaSparkContext(conf);
/**
* 通过编程入口加载数据
*/
JavaRDD<String> lines = sc.textFile("/Users/pengjunzhe/Downloads/spark-2.4.0-bin-hadoop2.7/README.md", 2);
/**
* 对数据进行处理
*/
JavaRDD<String> words = lines.flatMap(line -> Arrays.asList(line.split(" ")).iterator());
JavaPairRDD<String, Integer> ones = words
.mapToPair((PairFunction<String, String, Integer>) s -> new Tuple2<>(s, 1))
.reduceByKey((v1, v2) -> v1 + v2)
.filter(v1 -> v1._2 > 10)
.filter(v1 -> v1._1.length() > 1);
JavaPairRDD<Integer, String> transdOnes = ones
.mapToPair(t2 -> new Tuple2<>(t2._2, t2._1))
.sortByKey();
ones = transdOnes.mapToPair(t2 -> new Tuple2<>(t2._2, t2._1));
/**
* 对结果数据进行处理
*/
// ones.foreach(System.out::println);
ones.foreach(s -> System.out.println(s));
/**
* 结束关闭入口
*/
sc.stop();
}
}
注意最后不能用System.out::println的写法,抛出PrintStream不能被序列化的异常。
from pyspark import SparkConf, SparkContext
# 获取编程入口
conf = SparkConf() \
.setMaster("local[*]") \
.setAppName("WordCount")
sc = SparkContext(conf=conf)
# 通过编程入口加载数据
linesRdd = sc.textFile("file:///Users/pengjunzhe/Downloads/spark-2.4.0-bin-hadoop2.7/README.md")
# 对数据进行处理
wordCountRdd = linesRdd \
.flatMap(lambda x: x.split(" ")) \
.map(lambda x: (x, 1)) \
.reduceByKey(lambda x, y: x + y) \
.filter(lambda x: x[1] > 10) \
.filter(lambda x: x[0] != "" ) \
.sortBy(lambda x: x[1], False)
# 对结果数据进行处理
wordCountRdd.foreach(lambda x: print(x))
# 关闭编程
sc.stop()