CSE 344: Section 3

Grouping and (Un)Nesting

January 24th, 2019

1

Group By

• Powerful tool to handle “categories”
• Treat rows with a same attribute as a category
• Careful when selecting
• Only select attributes appeared in GROUP BY or aggregates
• SQLite will guess (arbitrarily pick a value)¯\_(ツ)_/¯
• SQL Server will throw an error ง •̀_•́)ง

2

Group By - Examples

Do these queries work?

​

Enrolled(stu_id, course_num)

​

​

​

​

SELECT stu_id, course_num

FROM Enrolled

GROUP BY stu_id

SELECT stu_id, count(course_num)

FROM Enrolled

GROUP BY stu_id

3

Group By - Examples

Do these queries work?

​

Enrolled(stu_id, course_num)

​

​

​

​

SELECT stu_id, course_num

FROM Enrolled

GROUP BY stu_id

SELECT stu_id, count(course_num)

FROM Enrolled

GROUP BY stu_id

4

Group By - Examples

Do these queries work?

​

Enrolled(stu_id, course_num)

​

​

​

​

SELECT stu_id, course_num

FROM Enrolled

GROUP BY stu_id

SELECT stu_id, count(course_num)

FROM Enrolled

GROUP BY stu_id

5

Witnessing (i.e. argmax)

Find the student who is taking the most classes.

Student(stu_id, id_num)

Enrolled(id_num, class)

​

SELECT S.stu_id

FROM Student S, Enrolled E

WHERE S.id_num = E.id_num

GROUP BY S.stu_id

HAVING COUNT(E.class) >= ALL(

SELECT COUNT(E1.class)

FROM Enrolled E1

GROUP BY E1.id_num);

6

Witnessing (i.e. argmax)

Find the student who is taking the most classes.

Student(stu_id, id_num)

Enrolled(id_num, class)

​

SELECT S.stu_id

FROM Student S, Enrolled E

WHERE S.id_num = E.id_num

GROUP BY S.stu_id

HAVING COUNT(E.class) = (

SELECT MAX(C) FROM (

SELECT COUNT(E1.class) AS C

FROM Enrolled E1

GROUP BY E1.id_num));

7

Alternative to “ALL”

Nested Queries

• Avoid when possible
• Danger of making simple queries slow and complicated
• Just because you can do it, doesn’t mean you should

8

Subquery in SELECT

SELECT DISTINCT C.cname, (SELECT count(*)

FROM Product P

WHERE P.cid=C.cid)

FROM Company C

​

​

​

9

Subquery in SELECT

Unnest using JOIN and GROUP BY

​

SELECT C.cname, count(P.cid)

FROM Company C LEFT OUTER JOIN

Product P ON C.cid = P.cid

GROUP BY C.cname;

10

Subquery in FROM

SELECT X.pname � FROM (SELECT *

FROM Product

WHERE price > 20) AS X

WHERE X.price < 500

​

​

More readable: WITH <alias> AS (<subquery>)

11

Subquery in FROM

Unnest using WHERE

​

SELECT X.pname � FROM Product AS X

WHERE X.price < 500 AND X.price > 20;

12

Subquery in WHERE

SELECT DISTINCT C.cname

FROM Company C

WHERE EXISTS (SELECT *� FROM Product P� WHERE C.cid = P.cid AND P.price < 200)

​

13

Subquery in WHERE

SELECT DISTINCT C.cname

FROM Company C, Product P

WHERE C.cid = P.cid AND P.price < 200

​

14

Subquery in WHERE Syntax

• SELECT ……… WHERE EXISTS (<sub>);
• SELECT ……… WHERE NOT EXISTS (<sub>);
• SELECT ……… WHERE attribute IN (<sub>);
• SELECT ……… WHERE attribute NOT IN (<sub>);
• SELECT ……… WHERE attribute > ANY (<sub>);
• SELECT ……… WHERE attribute > ALL (<sub>);

15

(Non-)monotonic Queries

• “Can we take back outputs by looking at more data?”
• Is this a monotonic query?

​

SELECT count(*)

FROM T1

GROUP BY T1.attr

16

(Non-)monotonic Queries

• “Can we take back outputs by looking at more data?”
• Is this a monotonic query?

​

SELECT count(*)

FROM T1

GROUP BY T1.attr

No! This query does not satisfy set containment.

​

Ex:

Current output: {(6), (23), (10)}

After more data: {(6), (23), (11)}

​

{(6), (23), (10)} ⊄ {(6), (23), (11)}

17

To Nest or Not to Nest

• Not an exact science
• Figuring out what is actually wanted will help you find simpler solutions (best way is to practice)
• Trigger words to use sub-querying
• Every, All (universal quantifiers)
• No, None, Never (negation)
• Only

18