CP6514 Lecture 4 Databases and GIS

1
UST/ENV/PAD/PDD 642/742 Introduction to
Geographic Information Systems
Lecture 4-1 GIS Database Management and
Thematic Mapping I
2
Attribute Database and GIS
The power of GIS lies not only in the pretty maps
it can make, but in the quality and amount of
attribute data behind the map
The referenced attribute data can be used to
drive thematic displays, perform spatial
analyses, and derive new data
3
Definition of a Database

• A Database is a logically coherent collection
  of data about some aspect(s) of the real world,
  with some inherent meaning
• The Database is designed, implemented and
  populated with data for a specific purpose
• More informally, a database is an organized
  collection of records (rows) representing
  distinct persons, places or things.
• Each record is comprised of a consistent set
  of fields (columns) which contains information
  describing that distinct person, place or thing

4
A Simple Database
5
Field Types Text

• The Text Field Type holds characters
• You can have a single character in a text
  string or a set of characters
• Depending on the type of information and the
  guiding business rules, values that consist
  entirely as digits should be considered strings
• In queries, text values should be enclosed in
  single or double quotes
• e.g. LU_CODE A
• Name Abigail Kristoffersson
• Zip_Code 30332

6
Field Types Integers

• Integer Field Types hold integer values
  restricted to particular integer ranges
• The Long Integer holds signed integers
  between 2,147,483,648 and 2,147,483,647
• The Short Integer holds signed integers
  between 32,768 and 32,767
• Integer arithmetic is very fast, but long
  integers have twice the memory requirements as
  short integers
• e.g. Age 32 short Int
• Yearned_Income 10000000 long Int

7
Field Types Floats and Doubles

• Floats and Doubles hold decimal numbers,
  depending on size and accuracy
• The Float Field Type holds decimal numbers
  with up to 7 significant digits
• The Double Field Type holds decimal numbers
  with up to 17 significant digits with accuracy
• Calculations are slower and approximate
• e.g. Height 1.720000 float
• X-Coord 2,134,587.916673144 double

8
Field Types Date

• The Date Field Type allows you to store both
  date and time information for a range between
  Jan. 1, 0100 and Dec. 1, 9999
• e.g. End_of_World Dec. 31, 1999 1159 PM
• End_of_World January 1, 2000
• End_of_World 1/1/2000 0001
• If no time is specified in the date field
  value, the system will assume 1200 midnight

9
Field Types BLOB

• BLOB stands for Binary Large Object and is the
  binary data type of a column in an RDBMS table
  that stores large image, text, or geometry data
• The BLOB field type essentially can store
  complex spatial information
• The BLOB field is often the RDBMS equivalent
  of the Shape field in the dBase component of a
  shapefile

10
Common Database Models

• Flat File Data Management Systems
• Hierarchical Data Management Systems
• Network Data Management Systems
• Relational Data Management Systems

11
Flat File Database and GIS
Single table of rows and columns Unique
identifier for each record High degree of
redundancy Several null values Not very
flexible
12
Relational Database Model

• Consists of several flat file tables or
  Entities
• Database defined both in terms of entities and
  their relationships
• Multiple tables can be joined to form a single
  virtual table
• Both entities and relationships are defined by
  unique identifiers or primary keys

13
JOINS

• In GISs, new virtual tables are usually
  formed when two or more tables are JOINED
• In JOINING operations, the source table fields
  are appended to the destination table, if the
  values in the specified join field coincide --
  you therefore have to specify a from source
  table, a source field item, a to destination
  table, and a destination field item before
  implementing a join
• JOINs are used when the relationship between a
  source and destination has a cardinality ratio
  of 1 1 or 1 N
• You can then make thematic displays or perform
  analyses using the fields of the source table
• You have the option of keeping all the source
  records, or keeping only those source records
  that had valid joins

14
One-to-One Relationship (JOIN)
TABLE 2 DESTINATION
TABLE 1 SOURCE
JOIN Relationship TABLE_1 . TR_NAME TABLE_2
. TRACT_NO
New Virtual Table
15
Many-to-One Relationship (JOIN)
TABLE 1 SOURCE
TABLE 2 DESTINATION
New Virtual Table
JOIN Relationship TABLE_1. Soil_Code
TABLE_2. Code
16
Advantages of Relational Model

• Easier Data Sharing
• Multi-user Transaction Processing
• Multiple Views of Database
• Flexibility, Scale Economies
• Access Control
• Redundancy Control
• Fewer Data Entry Errors
• Greater Data Consistency
• Ease in Backup and Recovery

17
Graphic and Attribute Data Linkages (JOIN)
Table 1
Table 2
(SOURCE)
(DESTINATION)
103
I-1
102
104
R-1
C-1
105
101
Interim Output from Table 1 and Table 2
Zoning Layer
Parcel Layer
Table 3.
(Source)
Final Output from Interim Output and Table 3
18
RELATE

• In RELATING operations, the procedure involves
  declaring the two tables that are going to be
  related, and the two fields that contain the
  relate values -- the relate operation then
  forms a relationship between the two tables,
  while visually keeping them separate
• RELATES can be used when the relationship
  between a source and destination has a
  cardinality ratio of N 1
• RELATES are primarily used to select related
  records in one table by selecting records in the
  other table
• Selections may be done either by location
  spatial query or by attribute query
• Each RELATE is given a user-specified name

19
RELATE (source destinationN1)
TABLE 1 SOURCE
TABLE 2 DESTINATION
20
Thematic Mapping Numeric Data Classification

• Used for Exploration purposes
• Used to make maps more readable, easier to
  understand
• Used to depict something about the data or area
  that is not self-evident

Numeric Data Classification Methods Natural
Breaks Standard Deviation Quantiles Equal
Interval User Defined or Manual
21
Natural Break Classification

• Data broken up into interval ranges such that
• Variance within each range is minimized
• Used for patterns/groupings inherent in data
• Uses Jenks optimization method

22
Quantile Classification

• Data broken up into interval ranges such that
• each range has an equal number of observations
• useful for linear patterns in data
• not very useful if disproportionate numbers of
  observations with very similar values occur

23
Equal Interval Classification

• Data broken up into intervals such that
• The range of each interval is equal

Manual or User-specified Classification

• Continuous - user sets UPPER LIMIT for each range
• Discontinuous - user sets both UPPER and LOWER
  LIMITS for each range similar to continuous,
  but with some ranges removed
• Discontinuous useful if no observations occur for
  some intermediate range -- these ranges may be
  skipped!

24
Standard Deviation Classification

• First, the mean is found for the entire data set
• Interval breaks found above and below the mean
  such that
• Breaks occur at 1/4 or 1/3 or 1/2 or 1 standard
  deviation(s) above and below the mean, till all
  the data is classified
• Useful in cases where the underlying population
  distribution is known
• Typically, a two-color ramp helps to emphasize
  values above in blue and below in red the
  mean

25
Classification by Normalization

• Instead of using the direct data value, a
  computation is made based on percent of total
• Alternately, one can compute a ratio of two data
  values for each record
• If the variation in the data values is extremely
  large, instead of using the direct data value, a
  computation is made based on the log value
• The Normalized Value then drives the data
  display

or
or