SQL Joins

v/s
Python Pandas
@vimanyuchaturvedi
INNER JOIN
LEFT_TABLE RIGHT_TABLE
ID LEFT_VALUE ID RIGHT_VALUE

1 LEFT 1 1 RIGHT 1 ID LEFT_VALUE RIGHT_VALUE

2 LEFT 2 4 RIGHT 2 1 LEFT 1 RIGHT 1

3 LEFT 3 5 RIGHT 3 4 LEFT 4 RIGHT 2

4 LEFT 4 6 RIGHT 4

SQL
SELECT * FROM LEFT_TABLE AS LT INNER JOIN RIGHT_TABLE AS RT
ON LT.ID = RT.ID

pandas
SELF JOIN
LEFT_TABLE LEFT_TABLE

ID LEFT_VALUE ID LEFT_VALUE

1 LEFT 1 1 LEFT 1

2 LEFT 2 2 LEFT 2

3 LEFT 3 3 LEFT 3

4 LEFT 4 4 LEFT 4

ID LEFT_VALUE LEFT_VALUE2

1 LEFT 1 LEFT 1

2 LEFT 2 LEFT 2

3 LEFT 3 LEFT 3

4 LEFT 4 LEFT 4

SQL
SELECT * FROM LEFT_TABLE AS LT INNER JOIN LEFT_TABLE AS LT2
ON LT.ID = LT2.ID

pandas
LEFT JOIN

LEFT_TABLE RIGHT_TABLE

ID LEFT_VALUE ID RIGHT_VALUE

1 LEFT 1 1 RIGHT 1

2 LEFT 2 4 RIGHT 2

3 LEFT 3 5 RIGHT 3

4 LEFT 4 6 RIGHT 4

ID LEFT_VALUE RIGHT_VALUE

1 LEFT 1 RIGHT 1

2 LEFT 2 NULL

3 LEFT 3 NULL

4 LEFT 4 RIGHT 2

SQL
SELECT * FROM LEFT_TABLE AS LT LEFT JOIN RIGHT_TABLE AS RT
ON LT.ID = RT.ID

pandas
RIGHT JOIN

LEFT_TABLE RIGHT_TABLE
ID LEFT_VALUE ID RIGHT_VALUE

1 LEFT 1 1 RIGHT 1

2 LEFT 2 4 RIGHT 2

3 LEFT 3 5 RIGHT 3

4 LEFT 4 6 RIGHT 4

ID LEFT_VALUE RIGHT_VALUE

1 LEFT 1 RIGHT 1

4 LEFT 4 RIGHT 2

5 NULL RIGHT 3

6 NULL RIGHT 4

SQL
SELECT * FROM LEFT_TABLE AS LT RIGHT JOIN RIGHT_TABLE AS RT
ON LT.ID = RT.ID

pandas
FULL JOIN
LEFT_TABLE RIGHT_TABLE
ID LEFT_VALUE ID RIGHT_VALUE

1 LEFT 1 1 RIGHT 1

2 LEFT 2 4 RIGHT 2

3 LEFT 3 5 RIGHT 3

4 LEFT 4 6 RIGHT 4

ID LEFT_VALUE RIGHT_VALUE

1 LEFT 1 RIGHT 1

2 LEFT 2 NULL

3 LEFT 3 NULL

4 LEFT 4 RIGHT 2

5 NULL RIGHT 3

6 NULL RIGHT 4

SQL
SELECT * FROM LEFT_TABLE AS LT FULL OUTER JOIN RIGHT_TABLE
AS RT ON LT.ID = RT.ID

pandas
UNION ALL
LEFT_TABLE RIGHT_TABLE
ID VALUE ID VALUE ID VALUE

1 VALUE 1 1 VALUE 1 1 VALUE 1

2 VALUE 2 4 VALUE 2 2 VALUE 2

3 VALUE 3 5 VALUE 3 3 VALUE 3

4 VALUE 4 6 VALUE 4 4 VALUE 4

1 VALUE 1

4 VALUE 2

5 VALUE 3

6 VALUE 4

SQL
SELECT * FROM LEFT_TABLE UNION ALL SELECT * FROM RIGHT_TABLE

pandas
UNION
LEFT_TABLE RIGHT_TABLE
ID VALUE ID VALUE

1 VALUE 1 1 VALUE 1

2 VALUE 2 4 VALUE 2

3 VALUE 3 5 VALUE 3

4 VALUE 4 6 VALUE 4

ID VALUE

1 VALUE 1

2 VALUE 2

3 VALUE 3

4 VALUE 4

4 VALUE 2

5 VALUE 3

6 VALUE 4

SQL
SELECT * FROM LEFT_TABLE UNION SELECT * FROM RIGHT_TABLE

pandas
INTERSECT
LEFT_TABLE RIGHT_TABLE

ID VALUE ID VALUE

1 VALUE 1 1 VALUE 1

2 VALUE 2 4 VALUE 2

3 VALUE 3 5 VALUE 3

4 VALUE 4 6 VALUE 4

ID VALUE

1 VALUE 1

SQL
SELECT * FROM LEFT_TABLE INTERSECT SELECT * FROM RIGHT_TABLE

pandas
EXCEPT
LEFT_TABLE RIGHT_TABLE

ID VALUE ID VALUE

1 VALUE 1 1 VALUE 1

2 VALUE 2 4 VALUE 2

3 VALUE 3 5 VALUE 3

4 VALUE 4 6 VALUE 4

ID VALUE

2 VALUE 2

3 VALUE 3

4 VALUE 4

SQL
SELECT * FROM LEFT_TABLE EXCEPT SELECT * FROM RIGHT_TABLE

pandas
SEMI JOIN
LEFT_TABLE RIGHT_TABLE

ID VALUE VALUE ID VALUE

VALUE 2 2 VALUE 2
1 VALUE 1
VALUE 3 3 VALUE 3
2 VALUE 2

3 VALUE 3

4 VALUE 4

SQL
SELECT * FROM LEFT_TABLE WHERE VALUE IN (SELECT VALUE FROM
RIGHT_TABLE )

pandas
ANTI JOIN
LEFT_TABLE RIGHT_TABLE
VALUE ID VALUE
ID VALUE

VALUE 2 1 VALUE 1
1 VALUE 1
VALUE 3 4 VALUE 4
2 VALUE 2

3 VALUE 3

4 VALUE 4

SQL
SELECT * FROM LEFT_TABLE WHERE VALUE NOT IN (SELECT VALUE
FROM RIGHT_TABLE )

pandas
CROSS JOIN

ID1 ID2
LEFT_TABLE RIGHT_TABLE
1 1
ID ID
1 2
1 1
2 1
2 2
2 2
3
3 1

3 2

SQL
SELECT * FROM LEFT_TABLE CROSS JOIN RIGHT_TABLE

pandas