Join是SQL語句中的常用操作，良好的表結構能夠將資料分散在不同的表中，使其符合某種正規化，減少表冗餘、更新容錯等。而建立表和表之間關係的最佳方式就是Join操作。

SparkSQL作為大資料領域的SQL實現，自然也對Join操作做了不少優化，今天主要看一下在SparkSQL中的Join，inner join，left outer join，right outer join，full outer join。

1.innerjoin：

程式碼實現：

1. // 啟動spark-shell，定義兩個rdd，做join操作

2. [[email protected] ~]$ spark-shell --master local[2]

3. scala> val a = sc.parallelize(Array(("A","a1"),("B","b1"),("C","c1"),("D","d1"),("E","e1"),("F","f1")))

4. a: org.apache.spark.rdd.RDD[(String, String)] = ParallelCollectionRDD[0] at parallelize at <console>:24

5. scala> val b = sc.parallelize(Array(("A","a2"),("B","b2"),("C","c1"),("C","c2"),("C","c3"),("E","e2")))

6. b: org.apache.spark.rdd.RDD[(String, String)] = ParallelCollectionRDD[1] at parallelize at <console>:24

7. scala> a.join(b).collect    // 這裡的join是inner join，只返回左右都匹配上的內容

8. res1: Array[(String, (String, String))] = Array((B,(b1,b2)), (A,(a1,a2)), (C,(c1,c1)), (C,(c1,c2)), (C,(c1,c3)), (E,(e1,e2)))

9. scala> b.join(a).collect    

10. res2: Array[(String, (String, String))] = Array((B,(b2,b1)), (A,(a2,a1)), (C,(c1,c1)), (C,(c2,c1)), (C,(c3,c1)), (E,(e2,e1)))

11. scala>

2.left outer join：

程式碼實現：

1. scala> a.leftOuterJoin(b).collect

2. res3: Array[(String, (String, Option[String]))] = Array((B,(b1,Some(b2))), (F,(f1,None)), (D,(d1,None)), (A,(a1,Some(a2))), (C,(c1,Some(c1))), (C,(c1,Some(c2))), (C,(c1,Some(c3))), (E,(e1,Some(e2))))

3. scala> b.leftOuterJoin(a).collect

4. res5: Array[(String, (String, Option[String]))] = Array((B,(b2,Some(b1))), (A,(a2,Some(a1))), (C,(c1,Some(c1))), (C,(c2,Some(c1))), (C,(c3,Some(c1))), (E,(e2,Some(e1))))

5. scala>

3.right outer join：

程式碼實現

1. scala> a.rightOuterJoin(b).collect

2. res4: Array[(String, (Option[String], String))] = Array((B,(Some(b1),b2)), (A,(Some(a1),a2)), (C,(Some(c1),c1)), (C,(Some(c1),c2)), (C,(Some(c1),c3)), (E,(Some(e1),e2)))

3. scala> b.rightOuterJoin(a).collect

4. res6: Array[(String, (Option[String], String))] = Array((B,(Some(b2),b1)), (F,(None,f1)), (D,(None,d1)), (A,(Some(a2),a1)), (C,(Some(c1),c1)), (C,(Some(c2),c1)), (C,(Some(c3),c1)), (E,(Some(e2),e1)))

5. scala>

4.full outer join：

程式碼實現：

1. scala> val a = sc.parallelize(Array(("A","a1"),("B","b1"),("C","c1"),("D","d1"),("E","e1"),("F","f1")))

2. a: org.apache.spark.rdd.RDD[(String, String)] = ParallelCollectionRDD[49] at parallelize at <console>:24

3. scala> val b = sc.parallelize(Array(("A","a2"),("B","b2"),("C","c1"),("C","c2"),("C","c3"),("E","e2")))

4. b: org.apache.spark.rdd.RDD[(String, String)] = ParallelCollectionRDD[50] at parallelize at <console>:24

5. scala> a.fullOuterJoin(b).collect

6. res15: Array[(String, (Option[String], Option[String]))] = Array((B,(Some(b1),Some(b2))), (F,(Some(f1),None)), (D,(Some(d1),None)), (A,(Some(a1),Some(a2))), (C,(Some(c1),Some(c1))), (C,(Some(c1),Some(c2))), (C,(Some(c1),Some(c3))), (E,(Some(e1),Some(e2))))

7. scala> b.fullOuterJoin(a).collect

8. res16: Array[(String, (Option[String], Option[String]))] = Array((B,(Some(b2),Some(b1))), (F,(None,Some(f1))), (D,(None,Some(d1))), (A,(Some(a2),Some(a1))), (C,(Some(c1),Some(c1))), (C,(Some(c2),Some(c1))), (C,(Some(c3),Some(c1))), (E,(Some(e2),Some(e1))))

9. scala>