The Relational Model and Normalization - Database Processing: Fundamentals, Design, and Implementation - page 124

EmployeeNumber FirstName LastName Department Email Phone Comment

100 Jerry Johnson Accounting [e-mail] 834-1101 Joined the

Accounting
Department in
March after
completing his
MBA. Will take the
CPA exam this fall.

200 Mary Abernathy Finance [e-mail] 834-2101

300 Liz Smathers Finance [e-mail] 834-2102

400 Tom Caruthers Accounting [e-mail] 834-1102

500 Tom Jackson Production [e-mail] 834-4101

600 Eleanore Caldera Legal [e-mail] 834-3101

700 Richard Bandalone Legal [e-mail] 834-3102 Is a full-time

consultant to Legal
on a retainer basis.

Figure 3-8
relation with Variable-
Length Column Values

that a relation has a certain column and contains 47 records. These three sets of terms are
summarized in Figure 3-9.

Functional Dependencies
Functional dependencies are the heart of the database design process, and it is vital for you to
understand them. We first explain the concept in general terms and then examine two exam-
ples. We begin with a short excursion into the world of algebra. Suppose you are buying boxes
of cookies and someone tells you that each box costs $5.00. With this fact, you can compute
the cost of several boxes with the formula:

CookieCost = NumberOfBoxes × $5

A more general way to express the relationship between CookieCost and NumberOfBoxes
is to say that CookieCost depends on NumberOfBoxes. Such a statement tells us the charac-
ter of the relationship between CookieCost and NumberOfBoxes, even though it doesn't give
us the formula. More formally, we can say that CookieCost is functionally dependent on
NumberOfBoxes. Such a statement can be written as:

NumberOfBoxes S CookieCost

This expression can be read as “NumberOfBoxes determines CookieCost.” The variable on the
left, here NumberOfBoxes, is called the determinant .

Figure 3-9
three Sets of Equivalent
terms

Table Column Row

Relation Attribute Tuple

File Field Record