repeat_clustering: Repeat a clustering configuration

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

Repeat a clustering configuration

Description

Repeat a clustering made with compare_clusterings() in order to obtain the TSClusters object.

Usage

repeat_clustering(series, clusterings, config_id, ...)

Arguments

series	The same time series that were given to compare_clusterings().
clusterings	The list returned by compare_clusterings().
config_id	The character indicating which configuration should be re-computed. Obtained from the clusterings' results' data frames.
...	More arguments for tsclust() (e.g. trace).

Details

Since the purpose of compare_clusterings() is to test many configurations, it is desirable to set its return.objects parameter to FALSE in order to save RAM. This function can then be used to compute the clustering object for a specific config_id.

Value

A TSClusters object.

Limitations

If the preprocessing function is subject to randomness, the clustering will not be correctly re-created by this function, since compare_clusterings() applies all preprocessing before calling tsclust().

If any parameters were given to compare_clusterings() through its ellipsis, they should probably be given to this function too.

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