Geography 353 Cartography and Visualization

...to Geog 353 Main Page and Course Description
...to Geog 353 Syllabus
...to Geog 353 Course Schedule and Lecture Outlines
...to Geog 353 Laboratory Information and Student Projects


Geog 353 Lecture Outline: Map Generalization and Classification
Update: 10/16/17

We will review chapter 8 (Map Generalization and Classification) from the Making Maps book. Additional information and examples can be gleaned from the material below.










Data Classification

Introduction

Recent Lectures: Issues concerning map symbolization: choosing visual marks to effectively represent the points, lines, and area data of our base maps and thematic data

Effective representation of intellectual hierarchy with a visual hierarchy

Visual variables

Next: ways to logically match the dimensions of your data (point, line, area) to symbols on your map

Requires understanding



1. Data Classification

Data is usually classified - put into some categories or groups - before it can be displayed

Different ways of classifying data will lead to different patterns on the map

Classification is a form of cartographic generalization which reduces the complexity of a set of thematic data

Classification: start by differentiating between




Categorical (qualitative, nominal) data classifications

Dealing with nominal (qualitative) data or data that is ordered but without a measurable range (rare as a type of mappable thematic data)


There are no absolute rules for this kind of classification, just general guidelines




Numerical data classifications


Ordered data with a measurable range: quantitative data

Two big issues involved in the classification of numerical or quantitative data




Number of classes:

Most maps for presentation purposes should have four to six classes

As you change the number of classes you may very well see different patterns:


Important to vary the number of classes and see what happens before you make a final choice

Number of Classes in ArcView




Number of Classes in ArcGIS








Also important is the way you divide up data: classification schemes


Data classification schemes



Histogram: graph relating data distribution and frequency

Some classification schemes take into account the distribution of the data, and others do not.


1. Exogenous schemes: class boundaries defined by criteria external to distribution of data


Advantage: map can be matched to external criteria

Disadvantage: does not take into account the data distribution


Exogenous schemes in ArcView


Exogenous schemes in ArcGIS





2. Arbitrary schemes: class boundaries are set by arbitrary criteria

Equal Intervals: class boundaries are defined by rounded numbers or regular divisions

Often chosen because the classification looks tidy

Simple to do by hand:


Advantages:

Disadvantage: not sensitive to the data distribution (if not rectangular)


Equal Interval schemes in ArcView


Equal Interval and Defined Interval schemes in ArcGIS





3. Ideographic schemes: class boundaries defined by the shape of the data distribution

Ideographic schemes take more effort because they are chosen based on some characteristics of the data distribution itself


3a. Natural Breaks: Attempt to find natural breaks in the data; classify data into groups that are somewhat distinct from each other. Can do this by hand using a cumulative frequency graph (or graphic array) and then look for natural breaks in the data and put class breaks at those points

A good default method: good to start with this and see if it works

How to do it: start by creating a histogram



Advantages:

Disadvantages:


Natural Breaks in ArcView


Natural Break (Jenks) scheme in ArcGIS





3b. Quantiles: puts an equal number of values in each class


Easy to calculate


Advantage:


Disadvantage:


Quantiles in ArcView


Quantile scheme in ArcGIS





4. Serial schemes: class boundaries are defined by statistical or mathematical functions


Standard Deviation


Class boundaries determined by the mean and standard deviation


Normal distribution: values near the mean occur more often

Other distributions: not normal: more dispersed


Standard deviation: a measure of how dispersed a set of data is



Advantages:

Disadvantages:


Standard Deviation schemes in ArcView


Standard Deviation scheme in ArcGIS





5. Unclassified Schemes


"Unclassed" choropleth maps: the number of categories is equal to the number of data values

Each value has a unique symbol


Advantages

Disadvantages


Unclassified Schemes in ArcView or ArcGIS





Sum: data classification

1) categorical (nominal, ordinal) vs numerical (interval, ratio) data


2) number of classes


3) dividing up data: numerical classification


Change the classification scheme or number of classes and you get a different map

If all three classification schemes are appropriate for the data distribution then select the classification scheme that best represents what you know about the actual data distribution.


E-mail: jbkrygier@owu.edu

...to Geog 353 Main Page and Course Description
...to krygier teaching page.
...to krygier top page.

OWU Home
OWU Geology and Geography Home