java spark 计算各个省份广告点击数的top3

数据格式：

时间戳 省份 城市 用户 广告

1589677806 河南 洛阳 user1 ad1
1589677807 河南 郑州 user1 ad1
1589677808 河南 洛阳 user2 ad1
1589677809 河南 洛阳 user3 ad2
1589677811 河南 郑州 user1 ad2
1589677813 河南 偃师 user1 ad2
1589677815 浙江 杭州 user1 ad1
1589677818 浙江 杭州 user2 ad1
1589677806 河南 郑州 user2 ad1

需求：

计算各个省份的广告点击排序top3

代码实现：

import org.apache.spark.SparkConf;
import org.apache.spark.api.java.JavaPairRDD;
import org.apache.spark.api.java.JavaRDD;
import org.apache.spark.api.java.JavaSparkContext;
import org.apache.spark.api.java.function.FlatMapFunction;
import org.apache.spark.api.java.function.Function;
import org.apache.spark.api.java.function.Function2;
import org.apache.spark.api.java.function.PairFunction;
import org.apache.spark.sql.sources.In;
import scala.Tuple2;

import java.util.*;

/**
 * # _*_ coding:utf-8 _*_
 * # Author:xiaoshubiao
 * # Time : 2020/5/17 9:13
 **/

public class anli_test {
    public static void main(String[] args) {
        SparkConf conf = new SparkConf().setMaster("local[*]").setAppName("test");
        JavaSparkContext sc = new JavaSparkContext(conf);
        JavaRDD<String> stringJavaRDD = sc.textFile("D:\\tmp\\rizhi.txt");
        JavaPairRDD<String, Integer> stringIntegerJavaPairRDD = stringJavaRDD.
                //转换成键值对的形式 如（省份-广告,点击数）
                mapToPair(
                        new PairFunction<String, String, Integer>() {
                            @Override
                            public Tuple2<String, Integer> call(String s) throws Exception {
                                String[] s1 = s.split(" ");
                                return new Tuple2<String, Integer>(s1[1] + "-" + s1[4], 1);
                            }
                        }
                )
                // 计算相同key的点击数和：（省份-广告,点击数和）
                .reduceByKey(new Function2<Integer, Integer, Integer>() {
                    @Override
                    public Integer call(Integer integer, Integer integer2) throws Exception {
                        return integer + integer2;
                    }
                });

        stringIntegerJavaPairRDD.
                //转换key的结构 （省份-广告,点击数和）=》(省份,(广告，点击数和))
                mapToPair(
                new PairFunction<Tuple2<String, Integer>, String, Tuple2>() {
                    @Override
                    public Tuple2<String, Tuple2> call(Tuple2<String, Integer> stringIntegerTuple2) throws Exception {
                        String[] s = stringIntegerTuple2._1().split("-");
                        Tuple2<String, Integer> stringIntegerTuple21 = new Tuple2<>(s[1], stringIntegerTuple2._2());

                        return new Tuple2<>(s[0],stringIntegerTuple21);
                    }
                }
        )
                // 按照k聚合
                .groupByKey()
                //对值进行排序
                .mapValues(
                new Function<Iterable<Tuple2>, Iterable>() {
                    @Override
                    public Iterable call(Iterable<Tuple2> tuple2s) throws Exception {
                        ArrayList<Tuple2<String, Integer>> tuple2s1 = new ArrayList<>();
                        Iterator<Tuple2> iterator = tuple2s.iterator();
                        while (iterator.hasNext()){
                            Tuple2 next = iterator.next();
                            tuple2s1.add(next);
                        }
                        tuple2s1.sort(
                                new Comparator<Tuple2<String, Integer>>() {
                                    @Override
                                    public int compare(Tuple2<String, Integer> o1, Tuple2<String, Integer> o2) {

                                        return o2._2() - o1._2();
                                    }
                                }
                        );
                        ArrayList<Tuple2<String,Integer>> t = new ArrayList<>();
                        Iterator<Tuple2<String, Integer>> iterator1 = tuple2s1.iterator();
                        Integer i = 0;
                        Integer n = 2;
                        while (iterator1.hasNext() & i<n){
                            t.add(iterator1.next());
                            i++;
                        }
                        return t;
                    }
                }
        )
                .collect().forEach(x->System.out.println(x));
    }


}