Vector Data Model

A more complex, but light weight format best suited for discrete objects.

Vector Data Model

Represents space discretely:

Objects delineated using sets of coordinate pairs.
Linked to descriptive attributes
Many attributes per object

Points

A point feature is an individual x, y coordinate pair representing a precise location.

“Zero-dimensional”
No length, width, or area

Points

Points are great for representing a variety of objects, depending on the scale:

Almost Any Scale

Trees
Stop signs
Fire hydrants

Points

Points are great for representing a variety of objects, depending on the scale:

Smaller Scales

Cities
Wild Fires
Airports

Lines

A line feature is a set of connected points. Must have a start and end point. May have middle points (vertices).

One-dimensional
Length
No width or area

Lines

Lines are also great for representing a variety of objects, depending on the scale:

Almost Any Scale

Hiking trails
Power lines
Water pipes

Lines

Lines are also great for representing a variety of objects, depending on the scale:

Smaller Scales

Roads
Rivers
Storm tracks

Polygons

A polygon feature consist of a set of three or more vertices connected by line segments (edges) that form an enclosed shape.

Two-dimensional
Length & width
Area

Polygons

Preferred for many objects depending on scale:

Almost Any Scale

Climate units
Lakes
Political boundaries

Polygons

Preferred for many objects depending on scale:

Larger Scales

Buildings
Roads
Cities

Interior Rings

All polygons are an enclosed shape. Some can also have interior rings (holes).

Each ring is a separate set of vertices and edges within the polygon
Interior rings cannot overlap

TopHat Question 1

The vector data model can represent objects as: (select all that apply)

Points
Lines
Polygons
Surfaces

Multi-part Objects

When an object has multiple parts, the vector model allows for:

Multi-polygons
Multi-lines
Multi-points

Resolution

Data resolution also applies to the vector model:

Less straightforward than for raster model
Spacing between vertices
- Smallest resolvable feature
Higher resolution = larger file size

TopHat Question 2

The resolution of a vector layer is defined by

The cell size
The spacing between vertices
The point size
The number of attributes

Attribute Tables

Non-spatial data is stored separate from spatial data.

Linked by an index
- A unique identifier
- No two objects can have the same index
- Often labelled as the “FID” in Arc

Attribute Tables

Non-spatial data is stored separate from spatial data.

Ability to store many attributes
Less redundancy than raster model
Easy to add new attributes

TopHat Question 3

In the vector data model, each object can only have one attribute. We must stack objects if we need multiple attributes.

True
False

Loss of Variability

Similar to the mixed pixel problem, feature is treated as homogenous

What a unit covers multiple values?
Any variability is lost and the unit is treated as homogenous

Vector Overlay

More computationally expensive than raster overlay.

Calculate the relative positions of ever vertex in every layer
- Determine overlap, containment, etc.
- Many operations depending task

Field Calculator

Fields can be “variables” in linear equations
- Won’t update if field values change
Not always intuitive
- Can’t calculate the sum of a column without extra steps

Field Calculator

Fields can be “variables” in linear equations
- Won’t update if field values change
Not always intuitive
- Can’t calculate the sum of a column without extra steps

Topology

The spatial relationships between geographic features.

Important for route finding
- Google maps uses topology for directions
- Rules by feature
  - Pedestrian
  - Car
  - Transit

Topology

The spatial relationships between geographic features.

We can set rules e.g.:
- Polygons (i.e. buildings) cannot overlap
- Points (i.e. fire hydrants) must be attached to lines (i.e. water mains)

Key Advantages

The vector data model is well suited for discrete objects:

The compact data structure allows for smaller file sizes relative to the raster dta model
Easy to query and select by attributes
- Topology > Proximity & Network Analysis
Graphic output is usually “cleaner”
- For final production maps, often useful to convert to vector format

TopHat Question 4

This term describes the spatial relationships between features:

Topology
Resolution
Vertex
Attributes
Vector

Shapefiles

One of the most common file types you will encounter, stores the coordinates of vertices plus metadata. Identified by a .shp tag.

Object type: points/multi-points, lines/multi-lines, or polygons/multi-polygons.
Only one type per .shp!
Coordinate reference system (CRS).
Attribute table.

GeoJSON

A simple, lightweight format for most commonly encountered in web mapping. Identified by a .json tag.

Unlike shapefiles, a GeoJSON can mix of geometries.
Encoded stylistic choices in the file.
Larger File Size
An Example)

Text Data

Table 1: BC Weather Stations
Name	Province	ClimateID	Latitude	Longitude
ACTIVE PASS	BC	1010066	48.87	-123.28
ALBERT HEAD	BC	1010235	48.40	-123.48
BAMBERTON OCEAN CEMENT	BC	1010595	48.58	-123.52
BEAR CREEK	BC	1010720	48.50	-124.00
BEAVER LAKE	BC	1010774	48.50	-123.35
BECHER BAY	BC	1010780	48.33	-123.63
BRENTWOOD BAY 2	BC	1010960	48.60	-123.47
BRENTWOOD CLARKE ROAD	BC	1010961	48.57	-123.45

TopHat Question 5

The Vector data model is almost always a better choice than the raster data model because:

It represents objects discretely
It has better resolution for a given file size
We can store many attributes for one object without creating redundancy
The vector data model is not always better than the raster data model