Chapter 2: Relational algebra

Additional Relational Operators

Lectured by : SEAK Leng

1 2011 - 2012
 Relational algebra
2

 Operation based query language

Operation
Basic operations:
•Selection
•Projection
•Cross Product
•Union
•Different
•Rename
Additional operations:
•Intersection
•Join
•Division

Complex query and query tree

 Relational algebra
3

 6 basic operators in relational algebra:

1. Select 𝜎 selects a subset of tuples from reln.
2. Project 𝜋 selects a subset of columns from reln.
3. Cartesian product × allows to combine two relations.
4. Set-difference − tuples in reln. 1, but not in reln2.
5. Union ∪ tuples in reln1 plus tuples in reln2.
6. Rename 𝜌 renames attribute(s) and relation.
 Relational algebra
4

 Additional operators in relational algebra:

1. Intersection ∩
2. Join ⋈<𝑗𝑜𝑖𝑛 𝑐𝑜𝑛𝑑𝑖𝑡𝑖𝑜𝑛>
3. National Join ⋈
4. Division ÷
 Some operators can be expressed in term of others.
 Set of relational algebra operation { 𝜎, 𝜋, ∪, 𝜌, −, ×} is complete.
Other relational algebra operations can be expressed as a sequence
of operations from this set.
 8. Assignment Operator
5

 Denoted by = or ⟵ . It is a binary operator.

 Assignment must always be made to a temporary relation variable.

 provides a convenient way to express complex queries.

Ex1: 𝜋FName, LName (𝜎𝐷𝑁𝑜=2 (𝐸𝑚𝑝𝑙𝑜𝑦𝑒𝑒)) Employee

ID FName LName Salary DNo
FName LName 1 Sok Dara 3000 3
Sam Sambath 2 Sam Sambath 4000 2
Sok Piseth 3 Sok Piseth 2500 2

TEMP ⟵ 𝜎𝐷𝑁𝑜=2 𝐸𝑚𝑝𝑙𝑜𝑦𝑒𝑒 R FirstName LastName

Sam Sambath
R(FirstName, LastName) ⟵ 𝜋𝐹𝑁𝑎𝑚𝑒,𝐿𝑁𝑎𝑚𝑒 (TEMP)
Sok Piseth
 8. Assignment Operator
6

 Sometimes easier to write expressions a piece at a time.

 Incrementaldevelopment
 Documentation of steps involved.

 Example2: in-line RA expression:

𝜋𝑆𝑡𝑁𝑎𝑚𝑒 (𝜎𝑆𝑇𝑈𝐷𝐸𝑁𝑇.𝑆𝑡𝐼𝑑=𝑏𝑜𝑟𝑟𝑜𝑤𝑠.𝑆𝑡𝐼𝑑 (𝜎𝑀𝑎𝑗𝑜𝑟="𝐶𝑆" (𝑆𝑇𝑈𝐷𝐸𝑁𝑇𝑆) × 𝑏𝑜𝑟𝑟𝑜𝑤𝑠))
 Equivalent sequence of operations by using temporary relation variable:

T1 ⟵ 𝜎𝑀𝑎𝑗𝑜𝑟="𝐶𝑆" 𝑆𝑇𝑈𝐷𝐸𝑁𝑇𝑆
𝑇2 ⟵ 𝜎𝑇1.𝑆𝑡𝐼𝑑=𝑏𝑜𝑟𝑟𝑜𝑤𝑠.𝑆𝑡𝐼𝑑 (𝑇1 × 𝑏𝑜𝑟𝑟𝑜𝑤𝑠)
Result ⟵ 𝜋𝑆𝑡𝑁𝑎𝑚𝑒 𝑇2
 1. Set-Intersection
7

 R ∩ S . It is binary operation. Relations must be union-compatible.

 Includes all tuples that are in both R and S.
 Derivation: R ∩ S = R – (R – S) or
R ∩ S = (R ∪ S) – ((R − S) ∪ (S − R))
STUDENT VOLUNTEER

FN LN FN LN FN LN

Ousa Ang Ousa Ang Ousa Ang

Sreylen Chab Sreylen Chab Sreylen Chab
Chhit Chharng
Pich Din
Sanwathnak Chhay

R S R∩S
 2. Join Operation
8

 Binary operator.
 Syntax: R ⋈<𝑗𝑜𝑖𝑛 𝑐𝑜𝑛𝑑𝑖𝑡𝑖𝑜𝑛> S
 Combines related tuples from 2 relations into a single relation.
 Like Cartesian Product, combine tuples from 2 relations into single
“longer” tuples, but only those that satisfy matching condition .
 Derivation: R ⋈<𝑗𝑜𝑖𝑛 𝑐𝑜𝑛𝑑𝑖𝑡𝑖𝑜𝑛> S = 𝜎<𝑗𝑜𝑖𝑛 𝑐𝑜𝑛𝑑𝑖𝑡𝑖𝑜𝑛> (𝑅 × 𝑆)
 R(a1,
a2, a3); S(b1, b2, b3)
R ⋈<𝑗𝑜𝑖𝑛 𝑐𝑜𝑛𝑑𝑖𝑡𝑖𝑜𝑛> S = T (a1, a2, a3, b1, b2, b3)
 Ex: Department ⋈𝑀𝑔𝑟_𝐼𝐷=𝐸𝐼𝐷 Employee
 2. Join Operation
9

 Ex: to retrieve the details of the manager of each department.

Employee EID FName LName DNo
Department DNo DName Mgr_ID
1 Sok Dara 3
1 GIC 4
2 Sam Sambath 2
2 GCA 2
3 Sok Piseth 2
4 Sokhan Nita 1

Dept_Mgr ← Department ⋈𝑀𝑔𝑟_𝐼𝐷=𝐸𝐼𝐷 Employee Temp ← Department × Employee

Dept_Mgr ← σMrg_ID=EID (Temp)
 Result of the join operation:
Dept_Mgr DNo DName Mgr_ID EID FName LName DNo

1 GIC 4 4 Sokhan Nita 1

2 GCA 2 2 Sam Sambath 2

 Theta Join Operation
10

 Condition Join or Theta-join or inner join a is attribute of R

 < 𝑗𝑜𝑖𝑛 𝑐𝑜𝑛𝑑𝑖𝑡𝑖𝑜𝑛 > : 𝑎𝜃𝑏 b is attribute of S
 Syntax: R ⋈𝑎𝜃𝑏 S 𝜃 ∈ {=, ≠, >, ≥, <, ≤}
 Ex:
R A B S C A B C

20 25 50 80 40 35

80 40 35
R ⋈𝐵>𝐶 S
 Practice
11

 Write the result from the following RA expression

1. 𝑟 ⋈𝐵<𝐷 𝑠
 Equi-join Operation
13

 If condition involves only the comparison operator “=”, the condition

join is also called Equi-Join.
Employee EID FName LName DNo
Department DNo DName Mgr_ID 1 Sok Dara 3
1 GIC 4 2 Sam Sambath 2
2 GCA 2 3 Sok Piseth 2
4 Sokhan Nita 1

Dept_Mgr ← Department ⋈𝑀𝑔𝑟.𝐼𝐷=𝐸𝐼𝐷 Employee

Dept_Mgr DNo DName Mgr_ID EID FName LName DNo

1 GIC 4 4 Sokhan Nita 1

2 GCA 2 2 Sam Sambath 2

 Practice
14

 Write the result from the following RA expression

2. 𝑟 ⋈𝐶=𝑆𝐶 𝜌𝑆 𝑆𝐶,𝐷 s
 Natural join
16

 Can be performed only if there is a common R A B C S C D

attribute in between the Relation. 80 40 101 101 P1

 Syntax: R ⋈ S or R∗S 30 45 101 102 P2

 No join condition. 89 50 102 103 P3

 Equi-join on attributes having identical names followed

by protection to remove duplicate attributes.
R⋈S
 Given the relations R(A, B, C, D) and S(B, D, E)
A B C D
1. National join can be applied if R ∩ 𝑆 ≠ ∅
80 40 101 P1
2. The result schema is (A, B, C , D, E)
30 45 101 P1
3. And the result of r ⋈ s is defined as:
89 50 102 P2
r ⋈ s = 𝜋𝑟.𝐴, 𝑟.𝐵, 𝑟.𝐶, 𝑟.𝐷, 𝑠.𝐸 𝜎𝑟.𝐵=𝑠.𝐵 ∧ 𝑟.𝐷=𝑠.𝐷 𝑟 × 𝑠
 Natural join
17

 Ex1: to retrieve the details of the department of each project.

PROJECT PID PName DNum DEPARTMENT DNum DName Mgr_ID
101 ProjectX 1 1 GIC 4
102 ProjectY 2 2 GCA 2
103 ProjectZ 2

Proj_Dept ← PROJECT ⋈ DEPARTMENT

Proj_Dept
PID PName DNum DName Mgr_ID
101 ProjectX 1 GIC 4
Result of the Natural join:
102 ProjectY 2 GCA 2
103 ProjectZ 2 GCA 2
 Natural join
18

 Ex2: to retrieve the details of the department of each project.

PROJECT PID PName DNum DEPARTMENT DNo DName Mgr_ID
101 ProjectX 1 1 GIC 4
102 ProjectY 2 2 GCA 2
103 ProjectZ 2

Proj_Dept ← PROJECT ⋈ 𝜌 𝐷𝑁𝑢𝑚, 𝐷𝑁𝑎𝑚𝑒, 𝑀𝑔𝑟_𝐼𝐷 (DEPARTMENT)

DEPT ← 𝜌 𝐷𝑁𝑢𝑚, 𝐷𝑁𝑎𝑚𝑒, 𝑀𝑔𝑟_𝐼𝐷 (DEPARTMENT)

Proj_Dept ← PROJECT ⋈ DEPT
 Natural join
19

 Ex2: to retrieve the details of the department of each project.

PROJECT PID PName DNum DEPARTMENT DNo DName Mgr_ID
101 ProjectX 1 1 GIC 4
102 ProjectY 2 2 GCA 2
103 ProjectZ 2

Proj_Dept ← PROJECT ⋈ 𝜌 𝐷𝑁𝑢𝑚, 𝐷𝑁𝑎𝑚𝑒, 𝑀𝑔𝑟_𝐼𝐷 (DEPARTMENT)

Proj_Dept
PID PName DNum DName Mgr_ID
101 ProjectX 1 GIC 4
Result of the Natural join:
102 ProjectY 2 GCA 2
103 ProjectZ 2 GCA 2
 Natural join
20

 Ex3: to retrieve the details of the department of the projectY.

PROJECT PID PName DNum DEPARTMENT DNum DName Mgr_ID
101 ProjectX 1 1 GIC 4
102 ProjectY 2 2 GCA 2
103 ProjectZ 2

Proj_Dept ← 𝜎𝑃𝑁𝑎𝑚𝑒="𝑃𝑟𝑜𝑗𝑒𝑐𝑡𝑌" (PROJECT) ⋈ DEPARTMENT

Proj_Dept
Result of the Natural join: PID PName DNum DName Mgr_ID
102 ProjectY 2 GCA 2
 Practice
21

 Write the result from the following RA expression

3. 𝑟 ⋈ 𝑠
 Practice: Write RA Expression
23

 1. List all students’ name and the name of the course they enroll in.
Course Student
CID Cname StudentID Sname

C1 Math 1 Chea Nary

C2 Physic 2 Sok Polin

3 Yi Sophea
StudentCourse
#StudentID #CID

1 C1

1 C2

2 C2

2 C1

3 C2
Practice: Write RA Expression
 Who has taken a course taught by Anderson? Find their name, course
number, semester and year.
 3. Division
28

 Syntax: 𝑅 ÷ 𝑆
 Precondition: attributes in S must be subset of attributes in R, 𝑆 ⊆ 𝑅 .
Let 𝑟, 𝑠 be relations on schemas R and S, respectively where
 R( A1, … Am , B1 , … Bn)

 S( B1 , … Bn)
 The result of 𝑟 ÷ 𝑠 is a relation on schema R – S = (A1, … Am)
 Suited for queries that includes the phrase “for all”.
 The result of the division operator 𝑟 ÷ 𝑠 consists of the set of tuples
from 𝑟 defined over the attributes R-S that match the combination of
every tuples in 𝑠 .
 3. Division
29

 Syntax: 𝑅 ÷ 𝑆 or 𝑅 ∕ 𝑆
 Suited for queries that includes the phrase “for all”.
 The result of the division operator 𝑟 ÷ 𝑠 consists of the set of tuples
from 𝑟 defined over the attributes R-S that match the combination of
every tuples in 𝑠 .
 Precondition: attributes in S must be subset of attributes in R, 𝑆 ⊆ 𝑅 .
Let 𝑟, 𝑠 be relations on schemas R and S, respectively where
 R( A1, … Am , B1 , … Bn);

 S( B1 , … Bn)
 The result of 𝑟 ÷ 𝑠 is a relation on schema R – S = (A1, … Am)
 3. Division
30

 Ex1:
 3. Division
31

 Division: 𝑅÷𝑆
= 𝜋𝑦 𝑅 − 𝜋𝑦 ((𝜋𝑦 (𝑅) × 𝑆) − 𝑅)
Where R(x, y) and S(y), y are attributes in 𝑅 and not is 𝑆.

T1 ⟵ 𝜋𝑦 R
T2 ⟵ 𝜋𝑦 ( T1 × S − R)
Result ⟵ T1 − T2
 3. Division
32

 Ex2: To retrieve the Employee ID (EID) of the employees working on all

projects.
EMPLOYEE EID PID PROJECT PID
Result EID

1001 1 1 1002

1002 1 2
1002 2
1003 2

Sol: Result ⟵ EMPLOYEE ÷ PROJECT

Or Result ⟵
πEID EMPLOYEE − πEID ((πEID (EMPLOYEE) × PROJECT) − EMPLOYEE)
 3. Division
33

 Ex2: To retrieve the Employee ID (EID) of the employees working on all

projects. 𝑦 T1 × S
EMPLOYEE EID PID PROJECT PID
T1 EID EID PID

R 1001 1 S 1 1001 1001 1

1002 1 2 1002 1001 2

1002 2 1003 1002 1

1003 2 (T1 × S ) − R T2 1002 2

EID 1003 1
EID PID
T1 ⟵ 𝜋𝑦 R 1003 2
1001 2 1001

T2 ⟵ 𝜋𝑦 ( T1 × S − R) 1003 1 1003 Result EID

1002
Result ⟵ T1 − T2
 3. Division
34

 Ex3: which students take all the courses ?

StudentCourse Course
#StudentID #CID CID Cname

1 C1 C1 Math

1 C2 C2 Physic

2 C2

2 C1

3 C2

StudentID

2
 3. Division
35

 Ex3: which students take all the courses ?

StudentCourse Course
#StudentID #CID CID Cname StudentID CID
1 C1
1 C1 C1 Math
1 C2
1 C2 C2 Physic
2 C1
2 C2
2 C2
2 C1
3 C1
3 C2
3 C2
𝑆𝑡𝑢𝑑𝑒𝑛𝑡𝐶𝑜𝑢𝑟𝑠𝑒 ÷ 𝜋𝐶𝐼𝐷 𝐶𝑜𝑢𝑟𝑠𝑒
StudentID = 𝜋𝑆𝑡𝑢𝑑𝑒𝑛𝑡𝐼𝐷 𝑆𝑡𝑢𝑑𝑒𝑛𝑡𝐶𝑜𝑢𝑟𝑠𝑒 − 𝜋𝑠𝑡𝑢𝑑𝑒𝑛𝑡𝐼𝐷 (
1 𝜋𝑠𝑡𝑢𝑑𝑒𝑛𝑡𝐼𝐷 𝑆𝑡𝑢𝑑𝑒𝑛𝑡𝐶𝑜𝑢𝑟𝑠𝑒 × 𝜋𝐶𝐼𝐷 𝐶𝑜𝑢𝑟𝑠𝑒 − 𝑆𝑡𝑢𝑑𝑒𝑛𝑡𝐶𝑜𝑢𝑟𝑠𝑒
2 )
 3. Division
36

 Practice: which employees work on all the critical projects?

 Reference
38

 R. Ramakrishnan et J. Gehrke Database Management Systems. 2003

(troisième édition)
 Jean-Marc Petit(2011). Fondamentaux de la modélisation des
données. Département Informatique-INSA de Lyon