== Physical Plan ==
VeloxColumnarToRow (51)
+- TakeOrderedAndProjectExecTransformer (50)
   +- ^ ProjectExecTransformer (48)
      +- ^ BroadcastHashJoinExecTransformer Inner BuildRight (47)
         :- ^ ProjectExecTransformer (43)
         :  +- ^ BroadcastHashJoinExecTransformer Inner BuildRight (42)
         :     :- ^ ProjectExecTransformer (35)
         :     :  +- ^ BroadcastHashJoinExecTransformer Inner BuildRight (34)
         :     :     :- ^ ProjectExecTransformer (18)
         :     :     :  +- ^ FilterExecTransformer (17)
         :     :     :     +- ^ WindowExecTransformer (16)
         :     :     :        +- ^ SortExecTransformer (15)
         :     :     :           +- ^ WindowGroupLimitExecTransformer (14)
         :     :     :              +- ^ FilterExecTransformer (13)
         :     :     :                 +- ^ ProjectExecTransformer (12)
         :     :     :                    +- ^ RegularHashAggregateExecTransformer (11)
         :     :     :                       +- ^ InputIteratorTransformer (10)
         :     :     :                          +- ColumnarExchange (8)
         :     :     :                             +- VeloxResizeBatches (7)
         :     :     :                                +- ^ ProjectExecTransformer (5)
         :     :     :                                   +- ^ FlushableHashAggregateExecTransformer (4)
         :     :     :                                      +- ^ ProjectExecTransformer (3)
         :     :     :                                         +- ^ FilterExecTransformer (2)
         :     :     :                                            +- ^ FileSourceScanExecTransformer parquet spark_catalog.default.store_sales (1)
         :     :     +- ^ InputIteratorTransformer (33)
         :     :        +- ColumnarBroadcastExchange (31)
         :     :           +- ^ ProjectExecTransformer (29)
         :     :              +- ^ FilterExecTransformer (28)
         :     :                 +- ^ WindowExecTransformer (27)
         :     :                    +- ^ SortExecTransformer (26)
         :     :                       +- ^ WindowGroupLimitExecTransformer (25)
         :     :                          +- ^ FilterExecTransformer (24)
         :     :                             +- ^ ProjectExecTransformer (23)
         :     :                                +- ^ RegularHashAggregateExecTransformer (22)
         :     :                                   +- ^ InputIteratorTransformer (21)
         :     :                                      +- ReusedExchange (19)
         :     +- ^ InputIteratorTransformer (41)
         :        +- ColumnarBroadcastExchange (39)
         :           +- ^ FilterExecTransformer (37)
         :              +- ^ FileSourceScanExecTransformer parquet spark_catalog.default.item (36)
         +- ^ InputIteratorTransformer (46)
            +- ReusedExchange (44)


(1) FileSourceScanExecTransformer parquet spark_catalog.default.store_sales
Output [4]: [ss_item_sk#1, ss_store_sk#2, ss_net_profit#3, ss_sold_date_sk#4]
Batched: true
Location: CatalogFileIndex [{warehouse_dir}/store_sales]
PushedFilters: [IsNotNull(ss_store_sk), EqualTo(ss_store_sk,4)]
ReadSchema: struct<ss_item_sk:int,ss_store_sk:int,ss_net_profit:decimal(7,2)>

(2) FilterExecTransformer
Input [4]: [ss_item_sk#1, ss_store_sk#2, ss_net_profit#3, ss_sold_date_sk#4]
Arguments: (isnotnull(ss_store_sk#2) AND (ss_store_sk#2 = 4))

(3) ProjectExecTransformer
Output [2]: [ss_item_sk#1, UnscaledValue(ss_net_profit#3) AS _pre_1#5]
Input [4]: [ss_item_sk#1, ss_store_sk#2, ss_net_profit#3, ss_sold_date_sk#4]

(4) FlushableHashAggregateExecTransformer
Input [2]: [ss_item_sk#1, _pre_1#5]
Keys [1]: [ss_item_sk#1]
Functions [1]: [partial_avg(_pre_1#5)]
Aggregate Attributes [2]: [sum#6, count#7]
Results [3]: [ss_item_sk#1, sum#8, count#9]

(5) ProjectExecTransformer
Output [4]: [hash(ss_item_sk#1, 42) AS hash_partition_key#10, ss_item_sk#1, sum#8, count#9]
Input [3]: [ss_item_sk#1, sum#8, count#9]

(6) WholeStageCodegenTransformer (3)
Input [4]: [hash_partition_key#10, ss_item_sk#1, sum#8, count#9]
Arguments: false

(7) VeloxResizeBatches
Input [4]: [hash_partition_key#10, ss_item_sk#1, sum#8, count#9]
Arguments: 1024, 2147483647, 10485760

(8) ColumnarExchange
Input [4]: [hash_partition_key#10, ss_item_sk#1, sum#8, count#9]
Arguments: hashpartitioning(ss_item_sk#1, 1), ENSURE_REQUIREMENTS, [ss_item_sk#1, sum#8, count#9], [plan_id=1], [shuffle_writer_type=hash]

(9) InputAdapter
Input [3]: [ss_item_sk#1, sum#8, count#9]

(10) InputIteratorTransformer
Input [3]: [ss_item_sk#1, sum#8, count#9]

(11) RegularHashAggregateExecTransformer
Input [3]: [ss_item_sk#1, sum#8, count#9]
Keys [1]: [ss_item_sk#1]
Functions [1]: [avg(UnscaledValue(ss_net_profit#3))]
Aggregate Attributes [1]: [avg(UnscaledValue(ss_net_profit#3))#11]
Results [2]: [ss_item_sk#1, avg(UnscaledValue(ss_net_profit#3))#11]

(12) ProjectExecTransformer
Output [2]: [ss_item_sk#1 AS item_sk#12, cast((avg(UnscaledValue(ss_net_profit#3))#11 / 100.0) as decimal(11,6)) AS rank_col#13]
Input [2]: [ss_item_sk#1, avg(UnscaledValue(ss_net_profit#3))#11]

(13) FilterExecTransformer
Input [2]: [item_sk#12, rank_col#13]
Arguments: (isnotnull(rank_col#13) AND (cast(rank_col#13 as decimal(13,7)) > (0.9 * Subquery scalar-subquery#14, [id=#2])))

(14) WindowGroupLimitExecTransformer
Input [2]: [item_sk#12, rank_col#13]
Arguments: [rank_col#13 ASC NULLS FIRST], rank(rank_col#13), 10, GlutenFinal

(15) SortExecTransformer
Input [2]: [item_sk#12, rank_col#13]
Arguments: [rank_col#13 ASC NULLS FIRST], false, 0

(16) WindowExecTransformer
Input [2]: [item_sk#12, rank_col#13]
Arguments: [rank(rank_col#13) windowspecdefinition(rank_col#13 ASC NULLS FIRST, specifiedwindowframe(RowFrame, unboundedpreceding$(), currentrow$())) AS rnk#15], [rank_col#13 ASC NULLS FIRST]

(17) FilterExecTransformer
Input [3]: [item_sk#12, rank_col#13, rnk#15]
Arguments: ((rnk#15 < 11) AND isnotnull(item_sk#12))

(18) ProjectExecTransformer
Output [2]: [item_sk#12, rnk#15]
Input [3]: [item_sk#12, rank_col#13, rnk#15]

(19) ReusedExchange [Reuses operator id: 8]
Output [3]: [ss_item_sk#16, sum#17, count#18]

(20) InputAdapter
Input [3]: [ss_item_sk#16, sum#17, count#18]

(21) InputIteratorTransformer
Input [3]: [ss_item_sk#16, sum#17, count#18]

(22) RegularHashAggregateExecTransformer
Input [3]: [ss_item_sk#16, sum#17, count#18]
Keys [1]: [ss_item_sk#16]
Functions [1]: [avg(UnscaledValue(ss_net_profit#19))]
Aggregate Attributes [1]: [avg(UnscaledValue(ss_net_profit#19))#20]
Results [2]: [ss_item_sk#16, avg(UnscaledValue(ss_net_profit#19))#20]

(23) ProjectExecTransformer
Output [2]: [ss_item_sk#16 AS item_sk#21, cast((avg(UnscaledValue(ss_net_profit#19))#20 / 100.0) as decimal(11,6)) AS rank_col#22]
Input [2]: [ss_item_sk#16, avg(UnscaledValue(ss_net_profit#19))#20]

(24) FilterExecTransformer
Input [2]: [item_sk#21, rank_col#22]
Arguments: (isnotnull(rank_col#22) AND (cast(rank_col#22 as decimal(13,7)) > (0.9 * ReusedSubquery Subquery scalar-subquery#14, [id=#2])))

(25) WindowGroupLimitExecTransformer
Input [2]: [item_sk#21, rank_col#22]
Arguments: [rank_col#22 DESC NULLS LAST], rank(rank_col#22), 10, GlutenFinal

(26) SortExecTransformer
Input [2]: [item_sk#21, rank_col#22]
Arguments: [rank_col#22 DESC NULLS LAST], false, 0

(27) WindowExecTransformer
Input [2]: [item_sk#21, rank_col#22]
Arguments: [rank(rank_col#22) windowspecdefinition(rank_col#22 DESC NULLS LAST, specifiedwindowframe(RowFrame, unboundedpreceding$(), currentrow$())) AS rnk#23], [rank_col#22 DESC NULLS LAST]

(28) FilterExecTransformer
Input [3]: [item_sk#21, rank_col#22, rnk#23]
Arguments: ((rnk#23 < 11) AND isnotnull(item_sk#21))

(29) ProjectExecTransformer
Output [2]: [item_sk#21, rnk#23]
Input [3]: [item_sk#21, rank_col#22, rnk#23]

(30) WholeStageCodegenTransformer (7)
Input [2]: [item_sk#21, rnk#23]
Arguments: false

(31) ColumnarBroadcastExchange
Input [2]: [item_sk#21, rnk#23]
Arguments: HashedRelationBroadcastMode(List(cast(input[1, int, false] as bigint)),false), [plan_id=3]

(32) InputAdapter
Input [2]: [item_sk#21, rnk#23]

(33) InputIteratorTransformer
Input [2]: [item_sk#21, rnk#23]

(34) BroadcastHashJoinExecTransformer
Left keys [1]: [rnk#15]
Right keys [1]: [rnk#23]
Join type: Inner
Join condition: None

(35) ProjectExecTransformer
Output [3]: [item_sk#12, rnk#15, item_sk#21]
Input [4]: [item_sk#12, rnk#15, item_sk#21, rnk#23]

(36) FileSourceScanExecTransformer parquet spark_catalog.default.item
Output [2]: [i_item_sk#24, i_product_name#25]
Batched: true
Location: InMemoryFileIndex [{warehouse_dir}/item]
PushedFilters: [IsNotNull(i_item_sk)]
ReadSchema: struct<i_item_sk:int,i_product_name:string>

(37) FilterExecTransformer
Input [2]: [i_item_sk#24, i_product_name#25]
Arguments: isnotnull(i_item_sk#24)

(38) WholeStageCodegenTransformer (8)
Input [2]: [i_item_sk#24, i_product_name#25]
Arguments: false

(39) ColumnarBroadcastExchange
Input [2]: [i_item_sk#24, i_product_name#25]
Arguments: HashedRelationBroadcastMode(List(cast(input[0, int, false] as bigint)),false), [plan_id=4]

(40) InputAdapter
Input [2]: [i_item_sk#24, i_product_name#25]

(41) InputIteratorTransformer
Input [2]: [i_item_sk#24, i_product_name#25]

(42) BroadcastHashJoinExecTransformer
Left keys [1]: [item_sk#12]
Right keys [1]: [i_item_sk#24]
Join type: Inner
Join condition: None

(43) ProjectExecTransformer
Output [3]: [rnk#15, item_sk#21, i_product_name#25]
Input [5]: [item_sk#12, rnk#15, item_sk#21, i_item_sk#24, i_product_name#25]

(44) ReusedExchange [Reuses operator id: 39]
Output [2]: [i_item_sk#26, i_product_name#27]

(45) InputAdapter
Input [2]: [i_item_sk#26, i_product_name#27]

(46) InputIteratorTransformer
Input [2]: [i_item_sk#26, i_product_name#27]

(47) BroadcastHashJoinExecTransformer
Left keys [1]: [item_sk#21]
Right keys [1]: [i_item_sk#26]
Join type: Inner
Join condition: None

(48) ProjectExecTransformer
Output [3]: [rnk#15, i_product_name#25 AS best_performing#28, i_product_name#27 AS worst_performing#29]
Input [5]: [rnk#15, item_sk#21, i_product_name#25, i_item_sk#26, i_product_name#27]

(49) WholeStageCodegenTransformer (10)
Input [3]: [rnk#15, best_performing#28, worst_performing#29]
Arguments: false

(50) TakeOrderedAndProjectExecTransformer
Input [3]: [rnk#15, best_performing#28, worst_performing#29]
Arguments: 100, [rnk#15 ASC NULLS FIRST], [rnk#15, best_performing#28, worst_performing#29], 0

(51) VeloxColumnarToRow
Input [3]: [rnk#15, best_performing#28, worst_performing#29]

===== Subqueries =====

Subquery:1 Hosting operator id = 13 Hosting Expression = Subquery scalar-subquery#14, [id=#2]
VeloxColumnarToRow (65)
+- ^ ProjectExecTransformer (63)
   +- ^ RegularHashAggregateExecTransformer (62)
      +- ^ InputIteratorTransformer (61)
         +- ColumnarExchange (59)
            +- VeloxResizeBatches (58)
               +- ^ ProjectExecTransformer (56)
                  +- ^ FlushableHashAggregateExecTransformer (55)
                     +- ^ ProjectExecTransformer (54)
                        +- ^ FilterExecTransformer (53)
                           +- ^ FileSourceScanExecTransformer parquet spark_catalog.default.store_sales (52)


(52) FileSourceScanExecTransformer parquet spark_catalog.default.store_sales
Output [4]: [ss_addr_sk#30, ss_store_sk#31, ss_net_profit#32, ss_sold_date_sk#33]
Batched: true
Location: CatalogFileIndex [{warehouse_dir}/store_sales]
PushedFilters: [IsNotNull(ss_store_sk), EqualTo(ss_store_sk,4), IsNull(ss_addr_sk)]
ReadSchema: struct<ss_addr_sk:int,ss_store_sk:int,ss_net_profit:decimal(7,2)>

(53) FilterExecTransformer
Input [4]: [ss_addr_sk#30, ss_store_sk#31, ss_net_profit#32, ss_sold_date_sk#33]
Arguments: ((isnotnull(ss_store_sk#31) AND (ss_store_sk#31 = 4)) AND isnull(ss_addr_sk#30))

(54) ProjectExecTransformer
Output [2]: [ss_store_sk#31, UnscaledValue(ss_net_profit#32) AS _pre_2#34]
Input [4]: [ss_addr_sk#30, ss_store_sk#31, ss_net_profit#32, ss_sold_date_sk#33]

(55) FlushableHashAggregateExecTransformer
Input [2]: [ss_store_sk#31, _pre_2#34]
Keys [1]: [ss_store_sk#31]
Functions [1]: [partial_avg(_pre_2#34)]
Aggregate Attributes [2]: [sum#35, count#36]
Results [3]: [ss_store_sk#31, sum#37, count#38]

(56) ProjectExecTransformer
Output [4]: [hash(ss_store_sk#31, 42) AS hash_partition_key#39, ss_store_sk#31, sum#37, count#38]
Input [3]: [ss_store_sk#31, sum#37, count#38]

(57) WholeStageCodegenTransformer (1)
Input [4]: [hash_partition_key#39, ss_store_sk#31, sum#37, count#38]
Arguments: false

(58) VeloxResizeBatches
Input [4]: [hash_partition_key#39, ss_store_sk#31, sum#37, count#38]
Arguments: 1024, 2147483647, 10485760

(59) ColumnarExchange
Input [4]: [hash_partition_key#39, ss_store_sk#31, sum#37, count#38]
Arguments: hashpartitioning(ss_store_sk#31, 1), ENSURE_REQUIREMENTS, [ss_store_sk#31, sum#37, count#38], [plan_id=5], [shuffle_writer_type=hash]

(60) InputAdapter
Input [3]: [ss_store_sk#31, sum#37, count#38]

(61) InputIteratorTransformer
Input [3]: [ss_store_sk#31, sum#37, count#38]

(62) RegularHashAggregateExecTransformer
Input [3]: [ss_store_sk#31, sum#37, count#38]
Keys [1]: [ss_store_sk#31]
Functions [1]: [avg(UnscaledValue(ss_net_profit#32))]
Aggregate Attributes [1]: [avg(UnscaledValue(ss_net_profit#32))#40]
Results [2]: [ss_store_sk#31, avg(UnscaledValue(ss_net_profit#32))#40]

(63) ProjectExecTransformer
Output [1]: [cast((avg(UnscaledValue(ss_net_profit#32))#40 / 100.0) as decimal(11,6)) AS rank_col#41]
Input [2]: [ss_store_sk#31, avg(UnscaledValue(ss_net_profit#32))#40]

(64) WholeStageCodegenTransformer (2)
Input [1]: [rank_col#41]
Arguments: false

(65) VeloxColumnarToRow
Input [1]: [rank_col#41]

Subquery:2 Hosting operator id = 24 Hosting Expression = ReusedSubquery Subquery scalar-subquery#14, [id=#2]


