查詢Hive中每個節點的子節點的數量

Question

我想要拿出最好的HiveQL查詢來獲取行列表，其中一列將具有該節點具有的（直接）子節點的數量。數據庫是分層的，所以它看起來是這樣的：

| ID | Some other column | ParentID | 
+-----------------------------------+ 
| 1 | XXXXXXXXXX x X X | NULL  | 
| 2 | XXXXXXXXXX x X X | 1  | 
| 3 | XXXXXXXXXX x X X | 2  | 
| 4 | XXXXXXXXXX x X X | 1  | 

而且我試圖把它查詢到的輸出是這樣的：

| ID | Some other column | child count | 
+--------------------------------------+ 
| 1 | XXXXXXXXXX x X X | 2   | 
| 2 | XXXXXXXXXX x X X | 1   | 
| 3 | XXXXXXXXXX x X X | 0   | 
| 4 | XXXXXXXXXX x X X | 0   | 

Answer 1

嘗試這樣的事情與LEFT JOIN。

SELECT a.id, 
     COALESCE (b.child_count, 0) "child count" 
FROM mytable a 
     LEFT JOIN (SELECT parentid, 
         Count(*) child_count 
        FROM mytable 
        GROUP BY parentid) b 
       ON a.id = b.parentid;