ANANT MISHRA AnantMishra30

## sparkJoins-broadcastjoin.py
from pyspark.sql.functions import broadcast

joinExpr = person["grad_section_id"] == graduateProgram["id"]

person.join(broadcast(graduateProgram), joinExpr).explain()

== Physical Plan ==
AdaptiveSparkPlan isFinalPlan=false
+- BroadcastHashJoin [grad_section_id#596L], [id#610L], Inner, BuildRight, false
   :- Project [_1#586L AS id#594L, _2#587 AS name#595, _3#588L AS grad_section_id#596L, _4#589 AS subject_enrolled_id#597]

## sparkJoins-sortMerge.py
joinExpr = person["grad_section_id"] == graduateProgram["id"]

person.join(graduateProgram, joinExpr).explain()

== Physical Plan ==
AdaptiveSparkPlan isFinalPlan=false
+- SortMergeJoin [grad_section_id#596L], [id#610L], Inner
   :- Sort [grad_section_id#596L ASC NULLS FIRST], false, 0
   :  +- Exchange hashpartitioning(grad_section_id#596L, 200), ENSURE_REQUIREMENTS, [id=#5805]
   :     +- Project [_1#586L AS id#594L, _2#587 AS name#595, _3#588L AS grad_section_id#596L, _4#589 AS subject_enrolled_id#597]

## sparkJoins-crossjoin.py
#1
joinType = "cross"
graduateProgram.join(person, joinExpression, joinType)

#2
person.join(graduateProgram)

#3
person.crossJoin(graduateProgram)

## sparkJoins-leftanti.py
joinType = "left_anti"

graduateProgram.join(person, joinExpression, joinType)

## sparkJoins-leftsemi.py
joinType = "left_semi"

graduateProgram.join(person, joinExpression, joinType)

## sparkJoins-rightouter.py
joinType = "right_outer"

person.join(graduateProgram, joinExpression, joinType)

## sparkJoins-leftouter.py
joinType = "left_outer"

person.join(graduateProgram, joinExpression, joinType)

## sparkJoins-outer.py
joinType = "outer"

person.join(graduateProgram, joinExpression, joinType)

## sparkJoins-functionDesc.py
DATAFRAME_1.join(DATAFRAME_2, JOIN_CONDITION, JOIN_TYPE)

# .join is function call on any pyspark dataframe
# DATFRAME_2 is a mandatory parameter
# JOIN_CONDITION is used to specify which key is used to match both dataframes
# JOIN_TYPE is used to specify which type of join to implement

## sparkJoins-inner.py
joinExpression = person["grad_section_id"] == graduateProgram["id"]

person.join(graduateProgram, joinExpression)
	from pyspark.sql.functions import broadcast

	joinExpr = person["grad_section_id"] == graduateProgram["id"]

	person.join(broadcast(graduateProgram), joinExpr).explain()

	== Physical Plan ==
	AdaptiveSparkPlan isFinalPlan=false
	+- BroadcastHashJoin [grad_section_id#596L], [id#610L], Inner, BuildRight, false
	:- Project [_1#586L AS id#594L, _2#587 AS name#595, _3#588L AS grad_section_id#596L, _4#589 AS subject_enrolled_id#597]
	joinExpr = person["grad_section_id"] == graduateProgram["id"]

	person.join(graduateProgram, joinExpr).explain()

	== Physical Plan ==
	AdaptiveSparkPlan isFinalPlan=false
	+- SortMergeJoin [grad_section_id#596L], [id#610L], Inner
	:- Sort [grad_section_id#596L ASC NULLS FIRST], false, 0
	: +- Exchange hashpartitioning(grad_section_id#596L, 200), ENSURE_REQUIREMENTS, [id=#5805]
	: +- Project [_1#586L AS id#594L, _2#587 AS name#595, _3#588L AS grad_section_id#596L, _4#589 AS subject_enrolled_id#597]
	#1
	joinType = "cross"
	graduateProgram.join(person, joinExpression, joinType)

	#2
	person.join(graduateProgram)

	#3
	person.crossJoin(graduateProgram)
	joinType = "left_anti"

	graduateProgram.join(person, joinExpression, joinType)
	joinType = "left_semi"

	graduateProgram.join(person, joinExpression, joinType)
	joinType = "right_outer"

	person.join(graduateProgram, joinExpression, joinType)
	joinType = "left_outer"

	person.join(graduateProgram, joinExpression, joinType)
	joinType = "outer"

	person.join(graduateProgram, joinExpression, joinType)
	DATAFRAME_1.join(DATAFRAME_2, JOIN_CONDITION, JOIN_TYPE)

	# .join is function call on any pyspark dataframe
	# DATFRAME_2 is a mandatory parameter
	# JOIN_CONDITION is used to specify which key is used to match both dataframes
	# JOIN_TYPE is used to specify which type of join to implement
	joinExpression = person["grad_section_id"] == graduateProgram["id"]

	person.join(graduateProgram, joinExpression)