或交叉连接的连接结果的条件

Question

我正在尝试在 spark 上加入两个数据集，我使用的是 spark 2.1 版，

SELECT * 
  FROM Tb1 
       INNER JOIN Tb2 
          ON Tb1.key1=Tb2.key1 
            OR Tb1.key2=Tb2.Key2

但它是交叉连接的结果，我怎样才能连接两个表并只获得匹配的记录？

我也试过左外连接，但它也迫使我改为交叉连接??

Answer 1

加入两次：

 select * 
   from Tb1 
      inner join Tb2 
         on Tb1.key1=Tb2.key1 
      inner join Tb2 as Tb22 
         on Tb1.key2=Tb22.Key2

或左加入两者：

 select * 
   from Tb1 
      left join Tb2 
         on Tb1.key1=Tb2.key1 
      left join Tb2 as Tb22 
         on Tb1.key2=Tb22.Key2

Answer 2

试试这个方法

from pyspark.sql import SQLContext as SQC

sqc = SQC(sc)
x = [(1,2,3), (4,5,6), (7,8,9), (10,11,12), (13,14,15)]
y = [(1,4,5), (4,5,6), (10,11,16),(34,23,31), (56,14,89)]
x_df = sqc.createDataFrame(x,["x","y","z"])
y_df = sqc.createDataFrame(y,["x","y","z"])

cond = [(x_df.x == y_df.x) | ( x_df.y == y_df.y)]

x_df.join(y_df,cond, "inner").show()

输出

+---+---+---+---+---+---+
|  x|  y|  z|  x|  y|  z|
+---+---+---+---+---+---+
|  1|  2|  3|  1|  4|  5|
|  4|  5|  6|  4|  5|  6|
| 10| 11| 12| 10| 11| 16|
| 13| 14| 15| 56| 14| 89|
+---+---+---+---+---+---+

或交叉连接的连接结果的条件

Or conditions on join result on cross join

sql

apache-spark

apache-spark-sql

pyspark

pyspark-sql