clusteringOutput-class {MLInterfaces} | R Documentation |
container for clustering outputs in uniform structure
Objects can be created by calls of the form new("clusteringOutput", ...)
.
partition
:"integer"
, labels for
observations as clustered silhouette
:"silhouette"
,
structure from Rousseeuw cluster package measuring cluster
membership strength per observationdistEnv
:"environment"
not in use prcomp
:"prcompObj"
a wrapped
instance of stats package prcomp output metric
:"character"
string
identifying the distance function used to orient objects in feature space call
:"call"
for auditing learnerSchema
:"learnerSchema"
, a
formal object indicating the package, function, and other attributes
of the clustering algorithm employed to generate this object RObject
:"ANY"
, the
unaltered output of the function called according to learnerSchema signature(x = "clusteringOutput")
: extract
the unaltered output of the R function or method called according
to learnerSchema signature(x = "clusteringOutput", y = "ANY")
:
a 4-panel plot showing features of the clustering, including the
scree plot for a principal components transformation and a display
of the partition in PC1xPC2 plane. For a clustering method that
does not have a native plot procedure, such as kmeans, the parameter
y should be bound to a data frame or matrix with feature data for all records;
an image plot of robust feature z-scores (z=(x-median(x))/mad(x)) and
the cluster indices is produced in the northwest panel. signature(object = "clusteringOutput")
: concise
report VJ Carey <stvjc@channing.harvard.edu>
showClass("clusteringOutput")