deldupMclust {edci} | R Documentation |
Delete clusters differing only by rounding errors or having maximization value zero.
deldupMclust(clust, prec=NULL, ncol=NULL, dz=TRUE)
clust |
numerical matrix, which columns contain the parameters of the clusters. |
prec |
optional positive integer. Number of decimal places for rounding. |
ncol |
Number of columns describing the clusters. |
dz |
optional boolean. With dz=TRUE clusters are deleted,
for which the objective function has value 0. |
Since clusters found by oregMclust
or circMclust
often
differ only by rounding errors, the function
deldupMclust
can be used for rounding and deleting
duplicates. If clust
has a column named count
their
values are summed appropriate. Otherwise such a column is added.
For parameter clust
the object returned from oregMclust
or circMclust
can be used. Alternatively, an arbitrary matrix
can be given, of which the first ncol
columns describe the
clusters.
The parameter prec
is the number of decimal places for
rounding; the default is no rounding.
With ncol
the number of columns, which describe the clusters,
can be given. This is not needed, if clust
is an object
returned from oregMclust
or circMclust
.
Return value is an object of the same type as clust
.
Tim Garlipp, garlipp@mathematik.uni-oldenburg.de
Müller, C.H., Garlipp, T. (2003) Simple consistent cluster methods based on redescending M-estimators with an application to edge identification in images, to appear in JMVA.