Scribd
Upload
77 views

Model

The document provides an overview of the relational model for a CS143 course. It discusses the key concepts including the data model, relational model, database construction steps, schema and instance, keys, null values, and SQL. The relational model represents data as relations (tables) with attributes (columns) and tuples (rows). Database construction involves domain analysis, design, table creation with DDL, loading data, and querying/updating with DML. SQL is used to interact with relational databases and define schemas, load data, and query/update tables.

Uploaded by

kevinazhu
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
77 views

Model

The document provides an overview of the relational model for a CS143 course. It discusses the key concepts including the data model, relational model, database construction steps, schema and instance, keys, null values, and SQL. The relational model represents data as relations (tables) with attributes (columns) and tuples (rows). Database construction involves domain analysis, design, table creation with DDL, loading data, and querying/updating with DML. SQL is used to interact with relational databases and define schemas, load data, and query/update tables.

Uploaded by

kevinazhu
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 7

CS143: Relational Model

Book Chapters
(4th) Chapters 1.3-5, 3.1, 4.11
(5th) Chapters 1.3-7, 2.1, 3.1-2, 4.1
(6th) Chapters 1.3-6, 2.105, 3.1-2, 4.5

Things to Learn
Data model
Relational model
Database construction steps

Data Model
The way we model/conceptualize/visualize/represent data
Need some representation to manage data in a computer
Many different ways to model data
Example (Airline flight): Graph model
Node: city
Edge: flight between cities
Label on edge: flight time, etc.
Example (Company hierarchy): Tree model
CEO Presidents Vice presidents Department heads . . .
Many others, like object model, XML model, network model, etc.
Models to learn in the class: Relational and E/R model

Example to Use in the Class


School information
Student(sid, name, age, GPA, address, . . . )
Class(dept, cnum, sec, title, instructor, . . . )
Enroll(sid, dept, cnum, sec)
...
1

Relational Model
Example: Student(sid, name, address, age, GPA)
Student
sid name
addr
age GPA
301 John
183 Westwood
19
2.1
303 Elaine 301 Wilshire
17
3.9
401 James 183 Westwood
17
3.5
208 Esther 421 Wilshire
20
3.1
All data is represented as relations (= tables)
Each relation has a set of attributes (= columns)
Each relation contains of a set of tuples (= rows)
Each attribute has a domain (= type)
Only atomic types
Similar to Excel spreadsheet

History of Relational Model


By far, the most significant invention in the history of DBMS
E.F. Codd, 1970
Completely revolutionized the field ($9B market in 2001 according to Gartner)
Before it, network and hierarchical model: difficult to use and pose queries
Turing Award, 1981
Extremely simple
Used by practically all commercial DBMS these days
An argument for simplicity
Initial research prototypes
IBM System R IBM DB2
Berkeley Postgres PostgresSQL
IBM and Oracle: first commercial RDB vendor
Still dominate the market

Concepts and Terminology


Schema
The structure of relations in database: relation name, attribute name, domain (optional).
Example:
Student(sid, name, address, GPA, age)
Course(dept: char(2), cnum: int, sec: int, unit: int, title: char(100))
char(2): string of length 2

Instance (= Data)
Actual contents (tuples) of relation (explain using the table example)
Schema Type, Instance Value
Schema Class, Instance Instance

Keys
A set of attributes that are known to be unique in the relation
Student(sid, name, address, GPA, age)
Course(dept, cnum, sec, unit, instructor, title)
Multiple keys possible
Course(dept, cnum, sec, unit, instructor, title)
Course(dept, cnum, sec, unit, instructor, title)
Course(dept, cnum, sec, unit, instructor, title)
Q: When do we need keys? How can they be used?

Name Scope
Names of relation: Unique across relations
Names of attributes: Unique in a table, same name in different tables OK

Set semantics
No duplicate tuples (different in SQL. More discussion later)
Q: Can a relation with no duplicates have no keys?
Tuple order does not matter
Authors of a paper: Need explicit ordering
Orders of attributes do not matter
3

Null value
Common interpretation
Do not know
Do not want to say
Not applicable
Example: Student(id, dept, name, GPA) before first quarter?
Complications from Null
Example: Student(id, dept, name, age, GPA)
Q1 : Find students whose age 20. Susans age is Null. Susan in the result?
Q2 : Find students whose age < 20. Susan in the result?
Q3 : Find students whose age 20 or age < 20 and. Susan in the result?
But Q3 = Q1 Q2 ? Something is wrong.
Relational algebra, SQL: 3-valued logic
Every condition is evaluated as True, False or Unknown
Various (arbitrary) rules to deal with anomalous situation
More discussion later
Unfortunately, Nulls are very common in DB.
Significant complexity in DBMS implementation

Steps in Database Construction


Flow chart diagram
1. Domain Analysis: Understand application-domain semantics being captured
E/R diagram
discussed in the latter half of the course
2. Database design: Design tables to capture the information
Relational design theory (funtional dependency, normal form, etc)
discussed in the latter half of the course
3. Table creation: using Database Definition Language
DDL: A language to define relations and their characteristics:
Schema, integrity constraints, indexes, . . .
4. Load: typically bulk-load. insert tuple possible
5. Query and update: using Data Manipulation Language
DML: A language to query and update relations
4

SQL and DDL, Load, DML


What is SQL?
Structured Query Language
The standard language for interacting with all commericial RDBMS
The history of SQL standard
SQL89 (Ansi SQL): first standard
SQL92 (SQL2): the main standard. several hundred pages
SQL3 (SQL99): no vendor supports it all! 1600 pages
SQL4 (SQL03): bug-fix release
SQL5 (SQL06): extensions for XML
SQL6 (SQL08): merge, truncate, diagnostics
We will mainly use the standard SQL92 in class. Individual product uses slight variations
of the standard. Some class query may not run on them.
SQL has many components
DDL: Schema definition, constraints, indexes, . . .
DML: data retrieval, modification, . . .
Transactions, Authorization, . . .
We learn schema definition part today.

Basic SQL Types


Basic SQL types (commonly used subset)
String
Char(n) padded fixed length
Varchar(n) variable length
Number
Integer 32bit
Decimal(5,2) 999.99
Real, Double 32bit, 64bit
Datetime
Date 2002-01-15
Time 13:50:00
Timestamp 2002-01-15 13:50:00 (On MySQL, Datetime is preferred)
Schema definition (table creation)
Course(dept, cnum, sec, unit, instructor, title)

CREATE TABLE Course (


dept CHAR(2) NOT NULL,
cnum INTEGER NOT NULL,
sec INTEGER NOT NULL,
unit INTEGER,
instructor VARCHAR(30),
title VARCHAR(30),
PRIMARY KEY(dept, cnum, sec) )
No Null in primary key

Course(dept, cnum, sec, unit, instructor, title)


Course(dept, cnum, sec, unit, instructor, title)
Course(dept, cnum, sec, unit, instructor, title)
CREATE TABLE Course (
dept CHAR(2) NOT NULL DEFAULT CS,
cnum INTEGER NOT NULL,
sec INTEGER NOT NULL,
unit INTEGER,
instructor VARCHAR(30),
title VARCHAR(30) DEFAULT,
PRIMARY KEY(dept, cnum, sec),
UNIQUE(dept, cnum, instructor),
UNIQUE(dept, sec, title) )
One primary key per table
Unique for other keys
Primary key, unique are enforced through index (more discussion later)
SQL92: No Null in primary key. Null OK for unique (DB2: No Null for unique).
DEFAULT for default values

SQL for dropping a table


DROP TABLE Course

Loading data
Vendor specific
Oracle, MySQL
LOAD DATA INFILE <datafile> INTO TABLE Course
DB2
IMPORT FROM <datafile> OF DEL INSERT INTO Course
Bulk-load from a comma separated file
6

Querying and update: Data Manipulation Language (DML)


DML: A language to query and update relations
Relational query languages
Relational algebra (formal), SQL (practical)
What is a relational query?
Input: relation
Output: relation
Input relation query Output relation
More details later

Things to remember
Data model
Schema
Instance
Relational model
relation, attribute, tuple, domain
key
null value
set semantics
Database construction steps
1. Domain analysis
2. Database design: E/R model, database design theory
3. Table creation: DDL
4. Load
5. Query and update: DML

You might also like