THE NEAREST NEIGHBOR ANALYSIS
INTRODUCTION
Geography is all about describing the spatial arrangement of features on the earth's
surface. Describing the nature of spatial distributions of phenomena is usually problematic and
this was done subjectively, thus its authenticity is questioned. When we talk about the nearest
neighbor analysis, we should know that it has its origin dated back to an attempt made by
botanist to provide quantitative descriptions of the distribution of trees or species. This attempt
was due to the pioneering work of Clark P.J and Evans F.C. Today, geographers have adopted
these methods to meet its requirement as a scientific discipline. Geographers are interested in
the study of the distribution pattern or locational pattern of phenomena or settlements in
space. There are many ways geographers analyse patterns of location one of which is the
nearest neighbor analysis. At this juncture, it becomes our onion to say what the nearest
neighbor analysis is all about.
MEANING OF NEAREST NEIGHBOR
Nearest neighbor analysis measure the linear distance between two or more specified
neighboring locations. We can apply nearest neighbour analysis to behavioral phenomena
which possess discrete spatial locations which may be mapped as points. Upton , G. and
Fingleton, B. (1985) asserted that nearest neighbor analysis is a method of exploring pattern in
locational data by comparing graphically the observed distribution of functions of
event-to-event or random point-to-event nearest neighbor distances, either with each other or
with those that may be theoretically expected from various hypothesized models, in particular
that of spatial randomness Nearest neighbor analysis utilizes the fundamental concept of
randomness. A distribution is random when each spatial unit in the area containing the points
has an equal opportunity of receiving a point. A non random point pattern is either more
clustered than random or more uniform than random. It is used for spatial geography (study of
landscapes, human settlement, CBDs, etc).
Nearest neighbor analysis reduces the simplifies spatial distributions of observed points
to a pattern description called random, more grouped than random (regular), or more clustered
than random. Relations between neighboring points are derived under the assumption that
such points are randomly distributed in terms of the Poisson distribution. These relations are
subsequently used to detect the presence of non randomness in given patterns. With the
Poisson distribution employed as a standard of comparison, a chi-square comparison has also
been employed (Thompson, 1956).
Although, we are interested in divergence from randomness along the R scale. We
denote nearest neighbour analysis as:
Rn = 2đ√n/A
Where;
Rn = nearest neighbor value describing the point pattern.
d= observed average neighbor distance.
n= total number of phenomena under study
A = area of the phenomena under study in kilometres square.
METHODOLOGY OF NEAREST NEIGHBOR ANALYSIS
¨ Delimit the study area where the geographer wants to carry out the task or the area of
interest to the geographer. This could be on the map or on the earth's surface.
¨ Calculate the mean distance (d) between phenomena. To find the mean distance, measure
the straight line distance between each point and its nearest neighbor. Divide the total of all
distance by the total number of points under study. The formula is : d =Σd/n
Where:
Σ = sigma means to add up the values of distance.
d = distance of phenomenon from the nearest neighbor.
n = total number of phenomenon under study.
¨ Next is to find the area of the maps or the squares in kilometres.
¨ Then calculate the nearest neighbor by substituting all the elements in the formula, ormula
with the values calculated above.
DESCRIPTION OF NEAREST NEIGHBOR ANALYSIS
Nearest neighbor analysis will produce values which ranges from 0 to 2.15. A value of 1.0
provides the standard of randomness, increasing R values are indicative of increasing
dispersion leading to a limiting case of regularity, and decreasing R values provide evidence of
increasing clustering (Clark and Evans, 1954; Dacey, 1960; Neft, 1966; Taylor, 1977). As I noted
earlier, the nearest neighbor statistic will produce values between 0 and 2.15 from which we
infer whether the distribution pattern is clustered, random or regular. I will now explain what I
mean by clustered, random or regular pattern.
Clustered Pattern : This occur when the pattern of distribution of settlement or phenomena is
closed to each other. In other words, the spatial arrangement of locations is closed to each
other without reasonable space between points or locations. This type of settlement or
arrangement usually occur in nodal regions such as road junctions, seaports or routes. Under
ideal condition a perfect clustered pattern will have a Rn of 0. However no perfect condition
exist anywhere in the world, consequently, we cannot have a nearest neighbor value of 0. Thus
an important point to make is that the nearest neighbor statistic cannot be 0 but must be less
than 1.
Random Pattern : A settlement is considered to be random when it does not follow a criteria
with which we can refer the distribution as cluster or regular. Here, we cannot say that pattern
or the arrangement is random neither can we say is regular. I can say that is combination of
clustered and regular pattern of distribution. The nearest neighbor value for a random
distribution of settlement or phenomena is 1.
Regular pattern : A settlement or locational arrangement of phenomena on the earth's surface
is spread to a reasonable amount of consideration. We can refer to the regular pattern as
dispersed pattern. The nearest neighbor statistic will produce a value of 2.15 under perfect
condition. But as I have said earlier, no perfect condition exist on the earth's surface. Thus,
when nearest neighbor statistic produce a value more than 1,we say it is tending towards
regular.
CONCEPTUAL AND METHODOLOGICAL PROBLEMS IN NEAREST NEIGHBOR
ANALYSIS
The nearest neighbor statistic is a useful tool for dealing with certain spatial
phenomena. Its successful application is determined, of course, on an understanding of its
conceptual and practical limitations cum the conditions under which its usefulness will be
maximized. These considerations are summarized in the following ways.
1. Distinguishing Pattern From Processes: The R scale provides a pattern tern description.
Patterns themselves are the product of underlying processes which develop over time
and space. Since patterns provide only static evidence of spacing, they must be
approached with notions derived from a rationalization of processes thought to evolve
over space. In this sense patterns are no more than abstractions derived by artificially
halting dynamic processes. Point patterns are therefore, a synthetic visual expression at
a given point in time of processes which continuously operate over space. For this
reason any observed point with a spatial distribution represents an event in time and
space (Taylor, 1977). Alternative values of R do not necessarily justify the conclusion that
either arandom or a systematic process is operating. A distribution of observed points
provides no direct information about the underlying processes giving rise to the observed
distribution (Amedeo and Golledge, 1975). Generalizations from patterns to processes
can be defended only when processes are considered in forming hypotheses concerning
changing R values which are observed over time (Getis, 1964; Pinder and Witherick, 1972;
Dawson, 1975).
2. Limiting Values of the R Scale: As points in space tend to cluster interpoint distances, d,
and hence R, would tend toward zero. The opposite tendency to clustering is dispersion.
Taken to the extreme, dispersion results in a limiting regular pattern based on a triangular
lattice and an R of 2.15. These theoretical limits to the R scale are rarely observed in
practice. The complex processes underlying spatial distributions tend to produce spatial
distributions which are more complicated than those reflected by the theoretical limits of
the R scale (Pinder and Witheric; 1972). Hence it has been observed (Taylor, 1977, p. 148)
that the empirical values of R are likely to range between 0.33 and 1.67.
3. Processes Generating Non random and Random Patterns: Extreme cases of clustering
and regularity are deterministic in the sense that they have singular R values associated
with unique average interpoint distance's which satisfy each extreme case. This
determinism does not hold for a randomly produced distribution. A number of random
patterns share an R value of 1. In fact, we may purposely use a random procedure to
generate a pattern whose R value differs from 1 (Dawson, 1975). In such a case a known
random process could be shown to produce a spatial pattern with a tendency toward
clustering. This is evidence of a well known phenomena in inferential statistics, namely
that variation in random processes must be taken into account in interpreting stochastic
or random events (Taylor, 1977). The R scale is not merely a descriptive tool which
measures a pattern. Various sub processes are capable of producing R values within the
central range of the R scale, resulting in a large number of R values which exhibit small
divergence from a random expectation. Still, statistical caution cannot replace an
understanding of the changes of processes which results in patterns. A narrow view of
the application of nearest neighbor analysis according to Dacey 1960 would state that
the test is limited to sensitivity to non randomness but cannot be employed to explore
hypotheses regarding uniform or concentrated spatial patterns. This view may
presuppose the existence of either random variables which produce random patterns or
variables which side by side contribute to clear cases of regularity or clustering. In a
broader view, Pinder and Witherick 1972 holds that the concept of randomness is not
satisfactory as datum. In the real world one might regard a given pattern as a deviation
from either or both extremes of regularity and clustering. In the case of diffusion, for
example, in certain settings residential sites may bear social meanings which reflect the
social leadership character of residents (e.g., larger corner lots). Such a situation could
contribute to a hierarchical regular adoption pattern over time. In pari-pasu, social
contagion would contribute to an observed clustering of adopters. The factors which
influence the location of adopters are unlikely to operate in a random fashion, but they
may distort extreme conditions of regularity or clustering so that the observed pattern is
matched by a random distribution. This cannot be attributed to the operation of random
forces, but instead might be seen as the significant complex interaction of location and
other factors.
4. Higher Order Analysis: Considering the space or distance between all points and their first
or closest neighbor can result in utilizing only a small portion of the information available
in a spatial pattern. Imagine pairs of points which are evenly dispersed over space. The
distance between any given point and its nearest neighbor would be small. The smaller
the R value the more clustered the spatial pattern. Unless each point is considered in
relation to all others, the character of the spatial distribution will be secluded by nearest
neighbors (Getis, 1964; Neft, 1966; Charlton, 1976; Vincent, 1976). The solution is to
identify and average the distance between each point and not only the original nearest
neighbor but succeeding neighboring points (second, third .... nth nearest neighbors). At
any order of analysis (so named to refer to the number of succeeding neighbors
considered) (Dacey, 1963) observed neighbor distances can be compared to expected
distances randomly generated from Poisson distributions. By definition, an R value of 1.0
will define a situation of randomness. The R value 2.15 indicates maximized regularity for
the spatial pattern.
5. Boundary Definition: one associated problem is that of a point whose nearest neighbor
lies outside a defined boundary. Ignoring the point results in a pattern description which
is biased toward dispersion. Including the point makes the definition of density subject to
challenge, but produces a less biased pattern examination. Boundary definition is a
conceptual rather than practical problem. Delineating the area of analysis in terms which
correspond to the conceptual nature of the problem studied will often eliminate the need
to consider outlying neighboring points. It is important that the spatial content which is
under study be logically justifiable.
6. Nature Of The Data
RESEARCH APPLICATIONS OF NEAREST NEIGHBOR ANALYSIS
Nearest neighbor analysis can be applied in research to determine various arrangements
of features (central place) which will be useful in making useful and informed decisions about
the environment. We can apply nearest neighbor analysis in:
1) Identifying the spatial nature of retail shopping patterns or points or settlements or other
phenomena.
2) We can apply it to study the concept of distance in spatial arrangement of phenomena.
3) The relationship between the perceived locations of retail facilities.
4) Features of cognitive maps representing customer images of the contents of large scale
shopping areas can be describe using nearest neighbor.
5) The nature of consumers' social interaction patterns.
6) The distribution of sets of ideal store locations.
CONCLUSION
Application of the nearest neighbor technique will increase our knowledge of its uses
and limitations and consequently, may enrich our understanding of human behavior as regard
to distance. Again it is a useful tool in Geographical studies in the sense that we are able to
study and describe spatial phenomena on the earth's surface as well as make a relatively
accurate generalisation on the basis of the interpreted data rather than making decisions or
inferences as regard to point pattern based on subject and manual methods.
REFERENCES
Amedeo D. and Golledge R. G. (1975): An Introduction To Scientific Reasoning In Geography.
Clark P. J and EVANS F. C., "Distance to Nearest Neighbor as a Measure of Spatial
Relationships in Populations,'' Ecology, 35 (October, 1954), 445-453.
Dacey M. F., "A Note on the Derivation of Nearest Neighbor Distances," Journal of Regional
Science, 2 (Summer 1960), 81-87.
Dacey M. F. and Tung T., "The Identification of Randomness in Point Patterns," Journal of
Regional Science, 4 (Summer 1962), 83-96.
Sanford L. Grossbart, Robert A. Mittelstaedt, and Gene W. Murdock (1978), "Nearest Neighbor
Analysis: Inferring Behavioral Processes From Spatial Patterns", in NA - Advances in Consumer
Research Volume 05, eds. Kent Hunt, Ann Abor, MI: Association for Consumer Research, Pages:
114-118.
Ajuebo Mildred (2014), Lecture notes on Topographical Map Analysis - Geo 213. (Unpublished).
Contacts :
Mr. Emmanuel Nelly Akamagune
Email :[email protected] ;[email protected]
Mobile phone number : +2348066876203
University of Benin, Benin-City, Nigeria.
Copyright: All rights reserved.
