- #[[CT230 - Database Systems I]]
- **Previous Topic:** [[Database System Introduction]]
- **Next Topic:** [[Introduction to SQL & DDL]]
- **Relevant Slides:** ![Lecture01.pdf](../assets/Lecture01_1663153689773_0.pdf)
-
- Why learn about relational DBMS?
	- 90% of industry /  enterprise / business applications are still relational DBMS or relational DBMS with extensions (e.g., OO Relational).
	- The majority of industry applications require:
		- **Correctness**
		- **Completeness**
		- **Efficiency** (Complex optimisation techniques & complex indexing structures).
	- Relational DBMS provide this.
-
- What is the **Relational Data Model**? #card
  card-last-interval:: 21.53
  card-repeats:: 4
  card-ease-factor:: 2.32
  card-next-schedule:: 2022-12-06T08:01:57.584Z
  card-last-reviewed:: 2022-11-14T20:01:57.584Z
  card-last-score:: 3
	- The **Relational Data Model** consists of collections of **relations** (often called *tables*) where each relation contains **tuples** (*rows*) and **attributes** (*columns*).
	- The relational data model is closely related to the file system model.
	- Relations are named.
- What is a **relation**? #card
  card-last-interval:: 31.36
  card-repeats:: 4
  card-ease-factor:: 2.8
  card-next-schedule:: 2022-11-20T16:24:50.257Z
  card-last-reviewed:: 2022-10-20T08:24:50.257Z
  card-last-score:: 5
	- A table.
- What are **attributes**? #card
  card-last-interval:: 100.92
  card-repeats:: 5
  card-ease-factor:: 2.9
  card-next-schedule:: 2023-02-23T18:22:23.155Z
  card-last-reviewed:: 2022-11-14T20:22:23.156Z
  card-last-score:: 5
	- **Attributes** are columns.
	- Columns / attributes are ^^almost always fixed^^ and do not change.
- What are **tuples**? #card
  card-last-interval:: 29.26
  card-repeats:: 4
  card-ease-factor:: 2.66
  card-next-schedule:: 2022-12-13T22:48:34.835Z
  card-last-reviewed:: 2022-11-14T16:48:34.835Z
  card-last-score:: 5
	- **Tuples** are rows.
	- Rows contain the data.
	- There is a variable number of rows.
- What is the **cardinality** of a relation? #card
  card-last-interval:: -1
  card-repeats:: 1
  card-ease-factor:: 2.8
  card-next-schedule:: 2022-11-15T00:00:00.000Z
  card-last-reviewed:: 2022-11-14T16:51:27.460Z
  card-last-score:: 1
	- The ^^number of tuples in a relation^^ is referred to as the **cardinality** of that relation.
-
- ## Attributes / Columns
	- Each attribute belongs to **one** *domain* and has a single:
		- Name
		- Data Type
		- Format
	- ### Naming Columns #card
	  card-last-interval:: 28.3
	  card-repeats:: 4
	  card-ease-factor:: 2.66
	  card-next-schedule:: 2022-11-17T15:34:15.662Z
	  card-last-reviewed:: 2022-10-20T08:34:15.663Z
	  card-last-score:: 5
		- Case is **not** significant in SQL.
		- No spaces allowed.
		- No reserved keywords (e.g., date) allowed.
		- Choose meaningful variable names.
		- If given the names of relations and attributes, use ^^exactly^^ what you are given.
	- ### Data Types
		- You must ^^specify the **data type**^^ of all attributes (columns) defined.
		- Common data types used include:
			- **varchar(N)**, where **N** is an integer - used for strings.
			- date
			- int
			- double
		- You often must specify the size - especially for integers and strings
		- ![image.png](../assets/image_1663057796785_0.png)
		- ### NULL
			- What are **null-valued attributes**? #card
			  card-last-interval:: 29.26
			  card-repeats:: 4
			  card-ease-factor:: 2.66
			  card-next-schedule:: 2022-12-13T22:48:31.350Z
			  card-last-reviewed:: 2022-11-14T16:48:31.350Z
			  card-last-score:: 5
				- **Null-valued attributes** are what occurs when the values of some attribute within a particular tuple may be unknown or may not apply to that particular tuple. A **null value** is used for these cases.
				- **NULL** is a special marker used in SQL to denote the ^^absence of a value.^^
			- In some cases, we wish to allow the possibility of a `NULL` value although they will often require extra handling (e.g., checking `if == NULL`).
			- In other cases, we want to prevent `NULL` from being entered as a value and specify `NOT NULL` as a **constraint** on data entry.
	- ### Atomic Attributes
		- What is an **atomic attribute**? #card
		  card-last-interval:: 33.64
		  card-repeats:: 4
		  card-ease-factor:: 2.9
		  card-next-schedule:: 2022-12-18T07:51:31.019Z
		  card-last-reviewed:: 2022-11-14T16:51:31.019Z
		  card-last-score:: 5
			- An **atomic attribute** is an attribute which contains a ^^single value of the appropriate type^^, generally meaning, "no repeating values of the same type".
			- The relational model should **only** have atomic values.
	- ### Composite Attributes
		- What is a **composite attribute**? #card
		  card-last-interval:: 33.64
		  card-repeats:: 4
		  card-ease-factor:: 2.9
		  card-next-schedule:: 2022-12-18T07:52:00.872Z
		  card-last-reviewed:: 2022-11-14T16:52:00.872Z
		  card-last-score:: 5
			- A **composite attribute** is an attribute that is composed of several atomic attributes.
				- E.g., `Name = FirstName, Middle Initial, Surname`.
			- We often want to decompose a composite attribute into atomic attributes unless there is a very good reason not to.
	- ### Multi-Valued Attributes
		- What is a **multi-valued attribute**? #card
		  card-last-interval:: 4.14
		  card-repeats:: 2
		  card-ease-factor:: 2.56
		  card-next-schedule:: 2022-11-26T16:37:36.766Z
		  card-last-reviewed:: 2022-11-22T13:37:36.767Z
		  card-last-score:: 5
			- A **multi-valued attribute** is an attribute which has lower and upper bounds on the number of values for an individual entry.
				- The ^^opposite of an atomic attribute.^^
			- The relational model should **not** store multi-valued attributes.
				- Database design / redesign should be used to deal with this issue by creating more attributes (columns) or more tables.
		- ### Derived Attributes
			- What are **derived attributes**? #card
			  card-last-interval:: 28.3
			  card-repeats:: 4
			  card-ease-factor:: 2.66
			  card-next-schedule:: 2022-12-12T23:52:22.646Z
			  card-last-reviewed:: 2022-11-14T16:52:22.646Z
			  card-last-score:: 3
				- A **derived attribute** is an attribute whose value can be determined from another attribute.
					- E.g., you can derive age from birthdate.
				- It is a good idea to not directly store attribute which can be derived from other attributes.
			-
- ## Collection of Relations
	- A Relational Data Model consists of a collection of relations (tables).
		- Tables are **cross-linked**.
-
	- A relational database usually contains many relations (tables) rather than storing all data in one single relation.
	- What is a **Relational Schema**? #card
	  card-last-interval:: 19.3
	  card-repeats:: 4
	  card-ease-factor:: 2.32
	  card-next-schedule:: 2022-12-04T03:03:56.818Z
	  card-last-reviewed:: 2022-11-14T20:03:56.818Z
	  card-last-score:: 3
		- A **relational schema**, $R$, is the ^^definition of a **table** in the database.^^ It can be denoted by listing the table name and the attributes:
			- $$R=\{A_1,A_2,...,A_n\}$$
				- where $A_i$ is an attribute.
			- E.g., with $n=3$, `works_on(essn, pno, hours)`.
- ## Linking Tables
	- Two ^^extremely important concepts^^ within the relational model which allows tables to be linked & cross-referenced are:
		- **Primary Key** attributes.
		- **Foreign Key** attributes.
	-
	- ### Primary Keys
		- Fundamental concept of **Primary Keys**: #card
		  card-last-interval:: 28.3
		  card-repeats:: 4
		  card-ease-factor:: 2.66
		  card-next-schedule:: 2022-11-17T15:34:03.291Z
		  card-last-reviewed:: 2022-10-20T08:34:03.291Z
		  card-last-score:: 5
			- ^^All tuples (rows) in a relation must be **distinct**.^^
			- To ensure this, we must have one or more attributes / columns whose data values will ^^always be unique for each tuple^^ - these attributes are called **key attributes** and are used to identify a tuple in the relation.
			- There may be a few possibilities for the **primary key** - these are called **Candidate Keys**.
				- One candidate key is ultimately chosen as the primary key during the Design Stage.
		- What is a **Primary Key**? #card
		  card-last-interval:: 23.43
		  card-repeats:: 4
		  card-ease-factor:: 2.42
		  card-next-schedule:: 2022-12-08T02:44:46.959Z
		  card-last-reviewed:: 2022-11-14T16:44:46.960Z
		  card-last-score:: 5
			- A **primary key** is defined as one or more attributes per table where:
				- there can only be one such primary key per table
				- the primary key can never contain the `NULL` value
				- all values entered for the primary key must be unique (no duplicates across the rows)
			- Often, primary keys are used as indices.
			- We use the convention (in writing) that attribute which form primary keys are $\text{\underline{underlined}}$.
	-
	- ### Foreign Keys
		- What is a **Foreign Key**? #card
		  card-last-interval:: 23.43
		  card-repeats:: 4
		  card-ease-factor:: 2.42
		  card-next-schedule:: 2022-12-08T02:52:18.585Z
		  card-last-reviewed:: 2022-11-14T16:52:18.585Z
		  card-last-score:: 3
			- A **Foreign Key** is an attribute, or a set of attributes, within one table that matches or **links to** a **candidate key** of some other table (possibly the same table).
			- More formally:
				- Given relations $r_1$ and $r_2$, a **foreign key** of $r_2$ is an attribute (or set of attributes) in $r_2$ where that attribute is a **candidate key** in $r_1$. Relations $r_1$ and $r_2$ may be the same relations.
		- #### Foreign Key Terminology
			- The **parent**, **master**, or **referenced** table is the relation containing the candidate key(s).
			- The **child** or **referencing** table / relation is the relation containing the foreign key.
-