Understanding

RDBMS
IT4GIS
Keith T. Weber, GISP
GIS Director
ISU-GIS Training and Research Center
FUNDAMENTALS
 RDBMS
• Relational Database Management System
• The “I” in GIS (Information)
 BTW
• The Data-to-Wisdom Pathway

Data Information Knowledge

Wisdom

cf.,
https://en.wikipedia.org/wiki/DIKW_pyramid
 Database software...
• Light Duty

• Medium Duty

• Heavy Duty
 Database software...

• Light Duty

• Medium Duty

• Heavy Duty
 Relational
Databases
• Why are databases
important?
– Drive many business
processes
– Store large amounts of
data
– Retrieve data quickly
Spreadsheets vs. Databases

• Integrit
y!
• Structu
re
RDBMS CONCEPTS AND TERMS
 Independence

• Physical
• Logical
 Logical Consistency Example
• Character based database design
– FirstName (1-4)
– LastName (5-10)
– Address (11-46)

• Record #1: Paul Bunyun, 100 Main Street,

Pocatello, ID 83201
NOTE: Record #2 starts at character #47
 What Happens When We Add a New
Field?
• New Field = ZIP+4 (47-50)
– Example, Paul’s ZIP+4 = 1234
• Scripts written and referring to the original design will
fail
• Record #1: Paul Bunyun, 100 Main Street, Pocatello,
ID 83201
• Record #2: 1234 Johnhe, nry 150 Main Street,
Pocatello, ID 8
 Integrity
• Important for consistency and
transaction management.
• Types:
– Domain: all values come from predefined
domains or are null
– Redundancy: problems can occur as a result
of repetitive storage that is not consistently
updated and from stored data that is derived
from other stored data. Redundant data
must be consistent.
 Integrity Types (cont’d)
– Constraint: Business integrity. Stored data must
not violate business rules.
– Entity: Every record must be uniquely
identifiable (index field or ObjectID)
– Referential: Relationships must not be
ambiguous. Two types…
Cascading or non-cascading
Enforcing Integrity Rules

• Programmatic
• Systematic
Key Fields and Index fields
• Unique Identifiers
• Relate fields
– Primary key
– Foreign key
RDBMS STRUCTURE
 Database Tables

Database

Table1 Table2 Table3

Table Structure
 Data Value Types
Type Name Storage Occupied/ data value Valid Domain Range

Short Integer 2 bytes -35768 to 32767

Long Integer 4 bytes -2147483648 to 2147483647

Float 4 bytes Any number from n-45 to n38

Double 8 bytes Any number from n-324 to n308

Text (string) 10 + max. length = bytes Any alphanumeric characters

Date 8 bytes Jan 1, 100 to Dec. 31 9999

LOB (variant) 22 + max. length = bytes Any alphanumeric characters

 (BTW) Raster Data Types Worth
•
Knowing
1_BIT—A 1-bit unsigned integer. The values can be 0 or 1.
• 2_BIT—A 2-bit unsigned integer. The values supported can be from 0 to 3.
• 4_BIT—A 4-bit unsigned integer. The values supported can be from 0 to 15.
• 8_BIT_UNSIGNED—An unsigned 8-bit data type. The values supported can be
from 0 to 255.
• 8_BIT_SIGNED—A signed 8-bit data type. The values supported can be from -
128 to 127.
• 16_BIT_UNSIGNED—A 16-bit unsigned data type. The values can range from 0
to 65,535.
• 16_BIT_SIGNED—A 16-bit signed data type. The values can range from -32,768
to 32,767.
• 32_BIT_UNSIGNED—A 32-bit unsigned data type. The values can range from 0
to 4,294,967,295.
• 32_BIT_SIGNED—A 32-bit signed data type. The values can range from -
2,147,483,648 to 2,147,483,647.
• 32_BIT_FLOAT—A 32-bit data type supporting decimals.
• 64_BIT—A 64-bit data type supporting decimals.
 Making Sense of all this…
• Recall, there are 8 bits in 1 byte
• Cross-reference
– 8-bit is byte data
– 16-bit is short integer (2 bytes)
– 32-bit (signed or unsigned) is long integer (4 bytes)
– 32-bit (float) is single-precision floating point (4 bytes)
– 64-bit is double-precision floating point (8 bytes)
DATABASE DESIGN
Basic Steps in Database
Design
• Understand and document the business’
needs.
– Problem statement
– Business object types
– Business relationships
– Business constraints
• Create an ERM
• Data and process inventory
• Develop tuple types
• Tuple types to tables
• Integrity
• Populate the database
 A Scenario...

• Develop a database for

the ISU Campus
 Document the business needs
• What problem or issue is this database going to
address?
• This is a business statement
– Improve student access
– Improve safety
– Improve efficiency
–…
READING A BUSINESS STATEMENT
 Identify Candidate Classes
• A candidate class may or may not remain a class
throughout the design process
– A candidate class may or may not become a table
• Try not to think about tables when reading the
business statement at this point
 Think Object-Oriented
• Classes are nouns
• A noun is a “person, places, and things”
 Methods
• Identifying candidate methods allows us
to better understand how the business
operates and how the Enterprise uses
GIS data.
• A method is a behavior…a relationship
between classes
• Ultimately, a connection between two
tables
• The candidate methods will describe an
inheritance, aggregation, or dependency
 And now…Verbs
• Candidate methods are verbs
– They show action
– They are behaviors
 Create an Entity Relationship Model
(ERM)
• Symbolized.
– Standard Representation
– Attribute Representation
– Entity Instance Representation

BUILDINGS
DINING
K BldgNum:
K Restaurant
126 Number
DINING
Name:
NameGraveley Hall
Type:
TypeEducation
of food
 Relationships
• Determine the Relationship between Entity
Types.
• Add these to the ERM

(More about database relationship classes later in

the semester)
 Data and Process Inventory
• Database
Dictionary • The name of the
– BldgName building

• Primary use of the

– Type building (e.g., 0 =
Unknown or n/a; 1 =
Education, 2 =
Offices, etc.)
– Floors • The total number of
floors
 Develop Tuple Types
• Use your ERM with relationships
• Perform a “Walk-through” exercise
– Simulate data is being added/retrieved in your
database.
• Create another ERM using Attribute
Representation
Tuple Types to Tables
ENTITY TYPES
TUPLE TYPES

TABLES

RELATIONSHIP TYPES
 Normalization

• First-Fifth Form Normal (1FN , 2FN ,…5FN)

• Academic
• Applied
 1FN

• “All values are atomic”

– Each cell in the table contains only a single data
value
• Eliminate repeating groups
– Puppy_Trick1, Puppy_Trick2, …
• Note: some tables will be OK as planned…
just check to make sure.
Check this (1FN)…
 2FN

• Satisfy 1FN and…

• Redundant data must be eliminated
– How?
– Example: Puppy_ID, Trick_ID, Trick_Name
Check this (2FN)…
 3FN

• Satisfy 1NF and 2FN and…

• No non-key attributes are dependent on
other non-key attributes.
– Example: Appointment_ID, Name, Date, Time,
Species
 After Normalization
• New tuple types will be created.
• New tables will be planned.
• Many-many relationships will be handled
using associative tables (bridge tables).
 De-Normalization
• What? Is this heresy?
 Designing the Actual RDBMS
• Visual modeling based upon your ERM and
Tuple type model.
• Implementation of integrity rules based upon
your business constraints.
 Populate the Database
• Questions and concerns to
revisit
– Null data
– Reporting discrepancies and
variations
– Measuring or estimating
methods
– Client utility/efficiency
The Last Step?
 Professional Hints and Tips
• Using Google drive, sharing files/folders and
communicating this in email
Questions?