1. 程式人生 > >sparksql的agg函式,作用:在整體DataFrame不分組聚合

sparksql的agg函式,作用:在整體DataFrame不分組聚合

1、 agg(expers:column*) 返回dataframe型別 ,同數學計算求值
df.agg(max("age"), avg("salary"))
df.groupBy().agg(max("age"), avg("salary"))
2、 agg(exprs: Map[String, String])  返回dataframe型別 ,同數學計算求值 map型別的
df.agg(Map("age" -> "max", "salary" -> "avg"))
df.groupBy().agg(Map("age" -> "max", "salary" -> "avg"))
3、 agg(aggExpr: (String, String), aggExprs: (String, String)*)  返回dataframe型別 ,同數學計算求值
df.agg(Map("age" -> "max", "salary" -> "avg"))
df.groupBy().agg(Map("age" -> "max", "salary" -> "avg"))
例子1:
scala> spark.version
res2: String = 2.0.2
 
scala> case class Test(bf: Int, df: Int, duration: Int, tel_date: Int)
defined class Test
 
scala> val df = Seq(Test(1,1,1,1), Test(1,1,2,2), Test(1,1,3,3), Test(2,2,3,3), Test(2,2,2,2), Test(2,2,1,1)).toDF
df: org.apache.spark.sql.DataFrame = [bf: int, df: int ... 2 more fields]
 
scala> df.show
+---+---+--------+--------+
| bf| df|duration|tel_date|
+---+---+--------+--------+
|  1|  1|       1|       1|
|  1|  1|       2|       2|
|  1|  1|       3|       3|
|  2|  2|       3|       3|
|  2|  2|       2|       2|
|  2|  2|       1|       1|
+---+---+--------+--------+
 
 
scala> df.groupBy("bf", "df").agg(("duration","sum"),("tel_date","min"),("tel_date","max")).show()
+---+---+-------------+-------------+-------------+
| bf| df|sum(duration)|min(tel_date)|max(tel_date)|
+---+---+-------------+-------------+-------------+
|  2|  2|            6|            1|            3|
|  1|  1|            6|            1|            3|

+---+---+-------------+-------------+-------------+
注意:此處df已經少了列duration和tel_date,只有groupby的key和agg中的欄位

例子2:
import pyspark.sql.functions as func
agg(func.max("event_time").alias("max_event_tm"),func.min("event_time").alias("min_event_tm"))