Introduction to Spatial Data Attributes
This section discusses how attribute data are stored in a GIS and issues of representing characteristics of features in vector and raster data structures. Databases are used to link spatial data to attribute data in a GIS.
DBMS - Database Management System
Entity/object - in GIS terminology this is a specific feature (usually in vector data structures) such as a single point representing a tree/well/city. In DBMS terminology this is a type of feature - e.g. roads, rivers, parcels. Each individual occurence of an object (again following Database terminology) is an instance. Make sure you understand the difference between how the GIS community uses the term object vs. how the Database community uses the term. I'll primarily use the GIS usage, the Bolstad book mostly uses the Database usage.
Record - Data are stored in relational tables that organize attribute data for objects. A record stores the attribute data for one feature object in a vector dataset. Records are also sometimes called tuples
key - used to relate tables together. In a GIS keys are used to link spatial and attribute data together, and to link attribute tables to other attribute tables.
primary key - for most GIS usages, a field/attribute used to uniquely identify an object (note slight deviation from defintion used on Bolstad). In a vector dataset a primary key must have one and only one instance of each value (e.g. parcel id's, county FIPS codes). An example of these are the two digit county codes used on Indiana license plates.
Example - database of landowners and databse of land parcels
Attributes in vector data structures vs. raster data structures
Look-up tables, one-to-one vs. one-to-many links
Joining attributes tables
Why are DBMS used?