pdc_configs: Helper function for preprocessing/distance/centroid...

In dtwclust: Time Series Clustering Along with Optimizations for the Dynamic Time Warping Distance

Helper function for preprocessing/distance/centroid configurations

Description

Create preprocessing, distance and centroid configurations for compare_clusterings_configs().

Usage

pdc_configs(
  type = c("preproc", "distance", "centroid"),
  ...,
  partitional = NULL,
  hierarchical = NULL,
  fuzzy = NULL,
  tadpole = NULL,
  share.config = c("p", "h", "f", "t")
)

Arguments

type	Which type of function is being targeted by this configuration.
...	Any number of named lists with functions and arguments that will be shared by all clusterings. See details.
partitional	A named list of lists with functions and arguments for partitional clusterings.
hierarchical	A named list of lists with functions and arguments for hierarchical clusterings.
fuzzy	A named list of lists with functions and arguments for fuzzy clusterings.
tadpole	A named list of lists with functions and arguments for TADPole clusterings.
share.config	A character vector specifying which clusterings should include the shared lists (the ones specified in ...). It must be any combination of (possibly abbreviated): partitional, hierarchical, fuzzy, tadpole.

Details

The named lists are interpreted in the following way: the name of the list will be considered to be a function name, and the elements of the list will be the possible parameters for the function. Each function must have at least an empty list. The parameters may be vectors that specify different values to be tested.

For preprocessing, the special name none signifies no preprocessing.

For centroids, the special name default leaves the centroid unspecified.

Please see the examples in compare_clusterings() to see how this is used.

Value

A list for each clustering, each of which includes a data frame with the computed configurations.

dtwclust documentation built on Sept. 11, 2024, 9:07 p.m.