algorithm argument in
adproclus() has been changed to ALS2 from
ALS1. This is because the ALS2 algorithm often
is less time-consuming with similar performance.New model selection tools for full dimensional ADPROCLUS and low
dimensional ADPROCLUS. mselect_adproclus() and
mselect_adproclus_low_dim() to estimate a range of
models.
New functions to plot the results of the model selection tools.
plot_scree_adpc() and
plot_scree_adpc_preselected().
Automatic model selection with
select_by_CHull().
Compute cluster means on arbitrary data with
cluster_means().
Characteristics of the baseline cluster Cl0 are now
also included in the summary output.
Documentation of plot_vars_by_comp() now includes
the correct argument vars_by_comp instead of
VarsByComp.
Give control over the number of decimal places to be displayed in
the summary output with the digits argument in
summary.adproclus() and
cluster_means().
Allow partial matches of arguments for plot choice in
plot().
Remove unnecessary arguments passed to the plotting functions
internally to give user the flexibility to supply them via
....
Correct description of the iterations argument in
adproclus() and adproclus_low_dim().
Set default title of plots to NULL.
plot_cluster_network() now also works when
clusters have an overlap of 1 observation.