Creates an evolving data stream generator from a list of Moving Generator Clusters (MGCs), each representing one cluster that can be static or change over time.
Usage
DSD_MG(dimension = 2, ..., labels = NULL, description = NULL)
add_cluster(x, c, label = NULL)
get_clusters(x)
remove_cluster(x, i)
# S3 method for class 'DSD_MG'
add_cluster(x, c, label = NULL)Arguments
- dimension
the dimension of the DSD object. All MGCs added to the stream need to have the same dimensionality.
- ...
initial set of MGCs.
- description
An optional string used by
print()to describe the data generator.- x
A
DSD_MGobject.- c
The cluster (MGC) that should be added to the
DSD_MGobject.- label, labels
integer representing the cluster label.
NArepresents noise. If labels are not specified, then each new cluster gets a new label.labelsis a vector with one label for each cluster in....- i
The index of the cluster that should be removed from the
DSD_MGobject. The index refers to the list returned byget_clusters().
Value
DSD_MG() returns a DSD_MG object (a DSD). add_cluster() and
remove_cluster() modify and invisibly return the stream, and
get_clusters() returns the list of MGC components.
Details
DSD_MG is able to generate complex data streams with known ground truth
that evolve over a period of time: clusters can move, grow, shrink, appear,
disappear, split or merge. The behavior of the stream is determined by its
components, the MGCs. See MGC for the available components, how
density and generator time work, and how to define cluster shapes.
Each point is returned with a .class column containing the label of the
component that produced it. By default, each component gets a new integer
label. Several components can share a label to form one ground-truth
cluster (e.g., to create non-convex clusters). Noise components
(MGC_Noise) get the label NA unless a label is specified.
The generator can be modified after it was created: add_cluster() adds a
component, remove_cluster() removes the component with a given index in
the list returned by get_clusters(). reset_stream() sets the generator
time back (to 1 by default). Note that streams and components are mutable
objects. Modifying a component also affects every stream that contains it.
See also
MGC for types of moving clusters.
Other DSD:
DSD(),
DSD_BarsAndGaussians(),
DSD_Benchmark(),
DSD_Cubes(),
DSD_Gaussians(),
DSD_Memory(),
DSD_Mixture(),
DSD_NULL(),
DSD_ReadDB(),
DSD_ReadStream(),
DSD_Target(),
DSD_UniformNoise(),
DSD_mlbenchData(),
DSD_mlbenchGenerator(),
DSF(),
animate_data(),
close_stream(),
get_points(),
plot.DSD(),
reset_stream()
Examples
### create an empty DSD_MG
stream <- DSD_MG(dimension = 2)
stream
#> Moving Data GeneratorClass: DSD_MG, DSD_R, DSD
#> With 0 clusters in 2 dimensions. Time is 1
### add two clusters
c1 <- MGC_Random(density = 50, center = c(50, 50), parameter = 1)
add_cluster(stream, c1)
stream
#> Moving Data GeneratorClass: DSD_MG, DSD_R, DSD
#> With 1 clusters in 2 dimensions. Time is 1
c2 <- MGC_Noise(density = 1, range = rbind(c(-20, 120), c(-20, 120)))
add_cluster(stream, c2)
stream
#> Moving Data GeneratorClass: DSD_MG, DSD_R, DSD
#> With 2 clusters in 2 dimensions. Time is 1
get_clusters(stream)
#> [[1]]
#> Random Moving Generator Cluster (MGC_Random, MGC)
#> In 2 dimensions
#>
#> [[2]]
#> Noise (MGC_Noise, MGC)
#> In 2 dimensions
#>
get_points(stream, n = 5)
#> X1 X2 .class
#> 1 50.35411 49.89896 1
#> 2 51.41467 49.49134 1
#> 3 49.97374 49.07762 1
#> 4 49.59217 48.48719 1
#> 5 50.25441 51.56740 1
plot(stream, xlim = c(-20,120), ylim = c(-20, 120))
if (interactive()) {
animate_data(stream, n = 5000, xlim = c(-20, 120), ylim = c(-20, 120))
}
### remove cluster 1
remove_cluster(stream, 1)
stream
#> Moving Data GeneratorClass: DSD_MG, DSD_R, DSD
#> With 1 clusters in 2 dimensions. Time is 10.902
get_clusters(stream)
#> [[1]]
#> Noise (MGC_Noise, MGC)
#> In 2 dimensions
#>
plot(stream, xlim = c(-20, 120), ylim = c(-20, 120))
### create a more complicated cluster structure (using 2 clusters with the same
### label to form an L shape)
stream <- DSD_MG(dimension = 2,
MGC_Static(density = 10, center = c(.5, .2), parameter = c(.4, .2),
shape = Shape_Block),
MGC_Static(density = 10, center = c(.6, .5), parameter = c(.2, .4),
shape = Shape_Block),
MGC_Static(density = 5, center = c(.39, .53), parameter = c(.16, .35),
shape = Shape_Block),
MGC_Noise( density = 1, range = rbind(c(0,1), c(0,1))),
labels = c(1, 1, 2, NA)
)
stream
#> Moving Data GeneratorClass: DSD_MG, DSD_R, DSD
#> With 4 clusters in 2 dimensions. Time is 1
plot(stream, xlim = c(0, 1), ylim = c(0, 1))
### simulate the clustering of a splitting cluster
c1 <- MGC_Linear(dimension = 2, keyframelist = list(
keyframe(time = 1, density = 20, center = c(0,0), parameter = 10),
keyframe(time = 50, density = 10, center = c(50,50), parameter = 10),
keyframe(time = 100,density = 10, center = c(50,100),parameter = 10)
))
### Note: The second cluster appears at time = 50
c2 <- MGC_Linear(dimension = 2, keyframelist = list(
keyframe(time = 50, density = 10, center = c(50,50), parameter = 10),
keyframe(time = 100,density = 10, center = c(100,50),parameter = 10)
))
stream <- DSD_MG(dimension = 2, c1, c2)
stream
#> Moving Data GeneratorClass: DSD_MG, DSD_R, DSD
#> With 2 clusters in 2 dimensions. Time is 1
dbstream <- DSC_DBSTREAM(r = 20, lambda = 0.1)
if (interactive()) {
purity <- animate_cluster(dbstream, stream, n = 2500, type = "micro",
xlim = c(-10, 120), ylim = c(-10, 120),
measure = "purity", horizon = 100)
}