Database: Introduction

Transcription

1 Database: Introduction Database Systems Important topic in the study of information systems Store/track items: scalar data (names, dates, ), pictures, audio, video Early applications: internal uses Internet: publishing database content to users Database design and development Understand, gather and organize user requirements Transform the designs into physical databases (applications) 1 2 What is a Database System? Description of System requirements An electronic filing cabinet A collection of data which h is: Integrated Collect/unify data from a number of different sources Shared Access by more than one person/application Advantages: Data is shared, redundancy can be reduced, and integrity maintained 3 Data analysis Look for entities, relationships and attributes ER Data model Data design Convert ER model (with cardinalities) into table definitions Logical Data model Physical Database bse design Choose appropriate p column types and constraints Physical Dt Data model dl 4

2 Elements of Entity-Relationship Model Entities, attributes, identifiers, and relationships Entity: Thing or object, whether real or imagined, about which information needs to be known/tracked Example: A customer called Peter Lam is an entity Entity can be Weak entity: existence-dependent on other entities Parent and child Strong entity: non-dependent on other entities Elements of Entity-Relationship Model: Entity Entity class/type: A collection that share CUSTOMER entity class contains: ti common properties or CustID characteristics FirstName UPPER CASE (noun) LastName Entity instance: phoneno Representation of a particular entity Two instances: Many instances in an entity class 1 Peter Lam Jenny Chan Elements of Entity-Relationship Model: Attributes Attributes: attribute Properties Value of the describe the entity s characteristics attribute phoneno CustID FirstName LastName phoneno An entity instance 1 Peter Lam Jenny Chan Elements of Entity-Relationship Model: Relationships Describes how one or more entities are related to each other Entity: noun, relationship: verb Binary Relationships (degree 2) Degree 3 relationships CUSTOMER EXAM CONTASS places attends determines ER Diagram ORDER SUBJECT GRADE 8

3 Types of Binary Relationship One-to-one one (1:1) One to many (1:N) Many to many (N:M) PROJECT GROUP PROJECT SUBJECT is_assigned_toi 1:1 is_assigned_to 1:N N:M enrolled 9 Relationship Maximum cardinality: The maximum number of entities that can occur on one side of the relationship 1:N, if a group project involves at most 3 students ts Maximum cardinality: 1:3 Minimum i cardinality: Indicates whether participation in the relationship is mandatory or optional 0: optional, 1: mandatory 10 Relationship is_assigned_toi PROJECT 1 1:1 0 Relationship is_assigned_to GROUP PROJECT 1 1N 1:N 0 entity optional Relationship Number entity involved One project May be Assigned dto One Student t One Student Must be Assigned one project entity optional Relationship Number involved Group project Student t entity Student project 11 12

4 SUBJECT Relationship 1 N:M 1 enrolled entity optional Relationship Number entity involved subject Student Student subject Attributes? 13 Example 1 Project: Project title (ProjTitle), Project ID (ProjID) and Budget (ProjBudget) Student: Year of study (YearStudy), Student Number (StudNo), First Name (FirstName) and Last Name (LastName) Relationship: A project might be assigned to one student A student must be assigned one project 14 Example 2 An order must contain at least one product. A product might not associate with any order An order consists of an order number as well as an order date The product name and its description will be listed out Also, the quantity of a particular product will be listed as well. Example 3 A subject must have at least one student enrolled A student must enroll at least one subject A subject has a title as well as a subject code A student has a student number, first name and last name The enrollment date should be recorded 15 16

5 Binary Relationships Entitles tables Subject table Two columns: subcode, title One-to-one one binary relationship Primary key on one side (mandatory side) becomes a foreign key on the other side (optional side) Both mandatory: place on one side only One side mandatory, one side optional: place on optional side 17 Design 1 Rule 1 is_assigned_to PROJECT 1:1 PROJECT ProjID title name studno ProjID firstname lastname Go from Student to Project use studno to obtain a student s row in obtain ProjID assigned to studno use this ProjID to look up the data in PROJECT 18 Design 2 Rule 1 PROJECT is_assigned_to 1:1 PROJECT ProjID title name studno studno firstname lastname Go from Project to Student use projid to obtain a project s row in PROJECT obtain studno assigned to projid use this studno to look up the student info in Rule 1 Both side mandatory: One project must assign to one student One student must be assigned one project Design 1 and design 2 are conceptually the same Might be different in performance Query in one direction is more common than query in the other direction One side mandatory, one side optional? A student must be assigned one project A particular project may not be assigned to any student Design 1: looks ok Design 2: null value Design 1 is preferable 19 20

6 Rule 2 One-to-many binary relationship GROUP PROJECT Key on parent: placed in child Cf: one-to-one relationship is_assigned_to GROUP PROJECT 1 1:N 0 ProjID title name studno ProjID firstname lastname Rule 3 many-to-many binary relationship Cannot be directly represented by relations as one-to-one or one-to-many cases Reason SUBJECT entity: {subcode, name} entity: {studno, firstname, lastname} Multiple values for Many-to-many relationship! SUBJECT 1 N:M 1 enrolled Rule 3 many-to-many binary relationship Create a new relation with primary keys of the two entities as its primary key Intersection relations Many-to-Many Relationship title Associative entities: SUBJECT enrolled 1 N:M 1 studno SUBJECT Subcode name studno firstname lastname SUBJECT- studno SubID code lastname dateenrolled firstname SUBJECT code title Natural? studno firstname lastname 23 SUBJECT- studno code dateenrolled 24

7 Many-to-Many Relationship Associative entities: SUBJECT code title studno firstname lastname SUBJECT- refno studno code dateenrolled Many-to-Many y Relationship Order: unique orderid and date Products: prodid, name, description Each order can include more than one product orderid orderdate ORDER contains 0 M:N 1 qty prodid PRODUCT prodname proddes Ternary Relationships CUSTOMER, SALESPERSON, ORDER Each order involves one and only one customer A customer can have many orders One order has one and only one sales A sales has many orders Ternary Relationship Further constraint? Each customer can place orders only with a particular salesperson Business rule Need to be documented 27 28

8 Relational Terminologies Pim Primary Key An attribute (or combination or attributes) that uniquely identifies a row in a relation Candidate Key Attribute that can be chosen to be the primary key Composite Key A primary key that t consists of more than one attribute Foreign Key An attribute in a relation that serves as the primary key of another relation Normalization Process for evaluating and converting a relation to reduce modification anomalies (undesirable consequence n of a data modification) Detects and eliminates data redundancy d Well structure relations should avoid Insertion anomaly Deletion anomaly Modification anomaly Example for Normal Forms Project Assignment: One project: assigned to a no of depts Unnormalized form: all information together ProjID Title DeptNo DeptName Locaton NoStaff HourChg ExChg Hours 123 A 12 Engineering China A 40 HRO HK A 70 Finance HK B 12 Engineering China B 70 Finance HK C 14 Purchase Taiwan D 70 Finance HK E 12 Engineering China E 70 Finance HK F 58 Sales China First Normal Form 1NF: A relation must have only single-valued attributes No repeating groups PROJECT entity: ProjID, DeptNo Title, DeptName, Location, NoStaff, HourChg, ExChg, Hours May have modification anomalies A new dept added? Delete ProjID 577? 32

9 1NF ProjID DeptNo Title DeptName Locaton NoStaff HourChg ExChg Hours A Engineering China A HRO HK A Finance HK B Engineering China B Finance HK C Purchase Taiwan D Finance HK E Engineering China E Finance HK F Sales China Key: {ProjID, DeptNo} 33 Second Normal Form 2NF: A relation must be in 1 NF and each non- key attribute must be dependent on the whole key To remove data which h is functionally dependent upon only part of a key and To place this data to a new relation, with that part of the key as the primary key 34 Second Normal Form PROJECT entity: Key: {ProjID, DeptNo} Title, DeptName, Location, NoStaffInv, HourChg, ExChg, Hours Dependency: ProjID Title DeptNo DeptName, Location, HourChg, ExChg {ProjID, DeptNo} NoStaffInv, Hours Split up Relations Second Normal Form: Answer PROJECT entity: ProjID, Title DEPT entity DeptNo, DeptName, Location, HourChg, ExChg ASSINMENT entity ProjID, DeptNo, NoStaffInv, Hours 35 36

10 ASSINMENT entity ProjID DeptNo NoStaff Hours NF PROJECT entity ProjID Title 123 A 234 B 345 C D E F DeptNo DeptName Locaton HourChg ExChg 12 Engineering China HRO HK 50 0 DEPT entity 70 Finance HK Purchase Taiwan Sales China Second Normal Form Problem: ExChg depends on Location If Location= HK, no ExChg If Location= China, ExChg = 20 If Location = Taiwan, ExChg = 40 Example: Delete deptno=14, deletion anomaly Third Normal Form 3NF: A relation must be in 2 NF and no transitive dependencies may exist A functional dependency between two or more non-key attributes DeptNo Location, ExChg Location ExChg ExChg depends on non-key attribute Location Solution: split relation Third Normal Form: Answer PROJECT entity: ProjID, Title DEPT entity DeptNo, DeptName, Location, HourChg EXTRACHARGE entity: y Location, ExChg ASSINMENT entity ProjID, DeptNo, NoStaffInv, Hours 39 40

11 ASSINMENT entity ProjID DeptNo NoStaff Hours NF EXTRACHARGE entity Location ExChg China 20 HK 0 Taiwan 40 PROJECT entity ProjID Title 123 A 234 B 345 C 456 D 567 E 577 F DeptNo DeptName Locaton HourChg 12 Engineering China HRO HK 50 DEPT entity 70 Finance HK Purchase Taiwan Sales China 60 Normalization vs Denormalization Normalization: Avoid modification anomalies Split relations Extra processing Might not be desirable Denormalization: control data redundancy to improve performance Example Assume one HoD, one to three Deputy HoD Un/De-normalized relation 1: DEPT (DeptName, HOD, DeputyHoD) Key: {DeptName, DeptyHoD} Not in : DeptName HoD Normalized DEPT (DeptName, HOD) DepHOD (DeptName, DeputyHOD) Example In 2NF: Not efficient: obtain data about dept, need to read at least 2 rows and as many as 4 rows Alternative DEPT (DeptName, HOD, DeHOD1, DeHOD2, DeHOD3) Key: DeptName In All attributes are functionally dependent on DeptName Problem: search the name of the DeHoD 43 44

12 SQL DBMS Access: Microsoft product Windows platform Easy to use: user-interface Small companies, personal uses MySQL server: Popular: available for windows, Linux, Unix Multiuser Multithread: Enable database to perform multiple tasks concurrently Increase server efficiency in handling client request Can handle large databases (tens of thousands of tables) (nearly) open source Graphical mode Graphical mode 47 48

13 MySQL Output format: table form Case insensitive i i Semi-colon: line ends 49 Data Definition Language Create database Create tables mysql> create database ABCpharmacy -> -> ; Query OK, 1 row affected (0.00 sec) mysql> use ABCpharmacy Database changed mysql> show tables; Empty set ( sec) mysql> DROP DATABASE db_name; 50 Data Definition Language Create tables Create table tablename (DEFINITION) DEFINITION: Col_name type [NOT NULL NULL] [PRIMARY KEY] INT: integer DOUBLE: floating-point number DECIMAL [M,D]: stores the digits in bytes: M: max display size, D: max decimals Char(M): fixed string: always occupying M bytes VARCHAR(M): variable length occupying at most M+1 bytes ENUM ( val1, val2, ) a string taken from a fixed set of values Color ENUM ( red, orange, yellow, green, blue, purple ) DESCRIBE tablename DROP TABLE tablename 51 52

14 DDL Text File Data Definition Language The text file must be tab-delimited Foreign key? Alter table table1name ADD FOREIGN KEY (columnname) REFERENCES table2name(columnname) Eg E.g., Alter table orderinfo add foreign key (custname) references customer (name); 55 56

15 Data Manipulation Language DML commands: Querying DB: SELECT SQL: querying tables Statements written in multiple lines Reserved words: in capital letters Conventions: provide clarity Do not required by compilers SELECT statement: retrieval SELECT <list of column expressions> FROM <list of tables and join operations> WHERE <list of logical expressions for rows> GROUP BY <list of grouping columns> HAVING <list of logical expressions for groups> ORDER BY <list of sorting specifications> Revision ER diagram Table design Normalization Denormalization n SQL Summary 59