Scribd
Upload
451 views3 pages

Nearest Neighbour Analysis

Nearest neighbor analysis measures the spatial distribution of points by calculating the distance between each point and its nearest neighbor. It provides a numerical value that indicates whether the points are clustered closely together or uniformly spaced apart. The analysis involves measuring distances between points categorized in the same group, like different types of buildings, and calculating a nearest neighbor index. This index can then be used to determine if the distribution of points is clustered, random, or uniform based on established scales. In an example, a researcher used nearest neighbor analysis to identify whether different land uses in a town, like retail or residential zones, were clustered or dispersed.

Uploaded by

Damian W
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
451 views3 pages

Nearest Neighbour Analysis

Nearest neighbor analysis measures the spatial distribution of points by calculating the distance between each point and its nearest neighbor. It provides a numerical value that indicates whether the points are clustered closely together or uniformly spaced apart. The analysis involves measuring distances between points categorized in the same group, like different types of buildings, and calculating a nearest neighbor index. This index can then be used to determine if the distribution of points is clustered, random, or uniform based on established scales. In an example, a researcher used nearest neighbor analysis to identify whether different land uses in a town, like retail or residential zones, were clustered or dispersed.

Uploaded by

Damian W
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

4i – A Guide to

Nearest Neighbour Analysis

Nearest Neighbour Analysis measures the spread or distribution of something over a


geographical space. It provides a numerical value that describes the extent to which a set of points
are clustered or uniformly spaced.

Why would we use nearest neighbour analysis?

Researchers use nearest neighbour analysis to determine whether the frequency with which
something is observed spatially is comparable with other locations. It can provide researchers with
a numerical value for the ‘clustering’ of a geographical phenomenon, allowing this value to be
compared more accurately with other places.

Worked Example:

In this example, researchers have mapped the land use of each building in a 200m by 200m plot in
a town centre, using a Goad map. A Goad map is a detailed street map that shows individual
buildings and their plots, and is usually needed for land registry and insurance purposes. They can
be purchased through online sources.

The nearest neighbour analysis can be used to identify whether there are clustered land use zones
within that section of the town.

This project was funded by the Nuffield Foundation, but the views expressed are those of the authors and not necessarily those of the Foundation.
The buildings have been categorised by their primary use: Retail; Financial; Municipal; Private
Other; Residential (Housing), and Green Space.

1. In each category, the buildings are coded and numbered. For example, in the retail
category the buildings are coded R1, R2, R3 etc.

2. For each building in each category, the distance to the next nearest building in that same
category is measured. Where measurements are taken between different points of the
map, as in this case, the researcher must ensure there is a high degree of consistency, by
always measuring from the centre of a feature or in the case of retail outlets, from the
centre of the main entrance doorway. This ensures that large buildings or areas are
subjected to exactly the same treatment as smaller ones. For ease in this example, only the
first ten buildings of the retail category will be analysed, but to cover the full land use of the
town, every category should be done in full.

Building R1 R2 R3 R4 R5 R6 R7 R8 R9 R10

Distance to nearest
8.2 4.5 6.9 9.6 9.5 12.3 23.4 18.5 5.0 12.9
neighbour (m)
3. The following calculation is used to work out the nearest neighbour value ( 𝑅n )

𝑅n = 𝐷(𝑂𝑂𝑂) where 𝐷(𝑂𝑂𝑂) = the mean value of the nearest neighbour


0.5 √ 𝑎 distances
𝑛 𝑎 = the area sampled
𝑛 = the number of points (in this case retail buildings)

𝑅n = 11.08 𝑅n = 11.08 𝑅n = 11.08 x


0.5 √ 40000 0.5 √ 4000 0.5 x 63.25
10

𝑅n = 11.08 𝑅n = 0.35
31.62

4. Using the following scale, the researcher can then identify the extent of clustering exhibited
by the category in question.

0 1 2.15

Clustered Random Uniform distribution

Therefore, the retail sector appears to be very clustered in the example town. However, the ten
points used in the example are too few to really show an accurate picture: the researcher should
aim for at least thirty to be able to say they have a significant result.

You might also like