Spark DataFrame的join操作

case class App(id: Int, name: String)
val data1 = sc.parallelize(Array((1, "a"), (2, "b"), (3, "c")))
val df1 = data.map { case (id, name) => App(id, name) }.toDF
val data2 = sc.parallelize(Array((1, "A"), (3, "B"), (5, "C")))
val df2 = data.map { case (id, name) => App(id, name) }.toDF

df1.show()
+---+----+
| id|name|
+---+----+
|  1|   a|
|  2|   b|
|  3|   c|
+---+----+

df2.show()
+---+----+
| id|name|
+---+----+
|  1|   A|
|  3|   B|
|  5|   C|
+---+----+

df1.join(df2)  // Cartesian join

df1.join(df2, "id")
+---+----+----+
| id|name|name|
+---+----+----+
|  1|   a|   A|
|  3|   c|   B|
+---+----+----+

df1.join(df2, df1("id") === df2("id")).show()
+---+----+---+----+
| id|name| id|name|
+---+----+---+----+
|  1|   a|  1|   A|
|  3|   c|  3|   B|
+---+----+---+----+

df1.join(df2, df1("id") === df2("id")).drop(df2("id")).show()
+---+----+----+
| id|name|name|
+---+----+----+
|  1|   a|   A|
|  3|   c|   B|
+---+----+----+