Please fine the query plan
scala> sqlContext.sql("SELECT dw.DAY_OF_WEEK, dw.HOUR, avg(dw.SDP_USAGE) AS
AVG_SDP_USAGE FROM (SELECT sdp.WID, DAY_OF_WEEK, HOUR, SUM(INTERVAL_VALUE)
AS SDP_USAGE FROM (SELECT * FROM date_d AS dd JOIN interval_f AS intf ON
intf.DATE_WID = dd.WID WHERE intf.DATE_WID >= 20150101 AND intf.DATE_WID <=
20150110 AND CAST(INTERVAL_END_TIME AS STRING) >= '2015-01-01 00:00:00.000'
AND CAST(INTERVAL_END_TIME AS STRING) <= '2015-01-10 00:00:00.000' AND
MEAS_WID = 3) AS test JOIN sdp_d AS sdp on test.SDP_WID = sdp.WID where
sdp.UDC_ID = 'SP-168451834' group by sdp.WID, DAY_OF_WEEK, HOUR) AS dw
group by dw.DAY_OF_WEEK, dw.HOUR")
q2: org.apache.spark.sql.SchemaRDD = SchemaRDD[36] at RDD at
SchemaRDD.scala:103
== Query Plan ==
== Physical Plan ==
Aggregate false, [DAY_OF_WEEK#3,HOUR#43L],
[DAY_OF_WEEK#3,HOUR#43L,(CAST(SUM(PartialSum#133), DoubleType) /
CAST(SUM(PartialCount#134L), DoubleType)) AS AVG_SDP_USAGE#126]
Exchange (HashPartitioning [DAY_OF_WEEK#3,HOUR#43L], 200)
Aggregate true, [DAY_OF_WEEK#3,HOUR#43L],
[DAY_OF_WEEK#3,HOUR#43L,COUNT(SDP_USAGE#130) AS
PartialCount#134L,SUM(SDP_USAGE#130) AS PartialSum#133]
Project [DAY_OF_WEEK#3,HOUR#43L,SDP_USAGE#130]
Aggregate false, [WID#49,DAY_OF_WEEK#3,HOUR#43L],
[WID#49,DAY_OF_WEEK#3,HOUR#43L,SUM(PartialSum#136) AS SDP_USAGE#130]
Exchange (HashPartitioning [WID#49,DAY_OF_WEEK#3,HOUR#43L], 200)
Aggregate true, [WID#49,DAY_OF_WEEK#3,HOUR#43L], [...
