pointdensity {dbscan} | R Documentation |
Calculate the local density at each data point as either the number of points in the eps-neighborhood (as used in DBSCAN) or the kernel density estimate (kde) of a uniform kernel. The function uses a kd-tree for fast fixed-radius nearest neighbor search.
pointdensity(x, eps, type = "frequency", search = "kdtree", bucketSize = 10, splitRule = "suggest", approx = 0)
x |
a data matrix. |
eps |
radius of the eps-neighborhood, i.e., bandwidth of the uniform kernel). |
type |
|
search, bucketSize, splitRule, approx |
algorithmic parameters for |
DBSCAN estimates the density around a point as the number of points in the
eps-neighborhood of the point (including the query point itself). The kde using a uniform
kernel is just this count divided by 2 eps n, where n is the number
of points in x
.
Points with low local density often indicate noise (see e.g., Wishart (1969) and Hartigan (1975)).
A vector of the same length as data points (rows) in x
with the
count or density values for each data point.
Michael Hahsler
WISHART, D. (1969), Mode Analysis: A Generalization of Nearest Neighbor which Reduces Chaining Effects, in Numerical Taxonomy, Ed., A.J. Cole, Academic Press, 282-311.
John A. Hartigan (1975), Clustering Algorithms, John Wiley \& Sons, Inc., New York, NY, USA.
frNN
.
set.seed(665544) n <- 100 x <- cbind( x=runif(10, 0, 5) + rnorm(n, sd=0.4), y=runif(10, 0, 5) + rnorm(n, sd=0.4) ) plot(x) ### calculate density d <- pointdensity(x, eps = .5, type = "density") ### density distribution summary(d) hist(d, breaks = 10) ### point size is proportional to Density plot(x, pch = 19, main = "Density (eps = .5)", cex = d*5) ### Wishart (1969) single link clustering method # 1. remove noise with low density f <- pointdensity(x, eps = .5, type = "frequency") x_nonoise <- x[f >= 5,] # 2. use single-linkage on the non-noise points hc <- hclust(dist(x_nonoise), method = "single") plot(x, pch = 19, cex = .5) points(x_nonoise, pch = 19, col= cutree(hc, k = 4)+1L)